Skip to main content

Full text of "A history of English sounds from the earliest period, with full word-lists"

See other formats


Google 



This is a digital copy of a book that was preserved for generations on library shelves before it was carefully scanned by Google as part of a project 

to make the world's books discoverable online. 

It has survived long enough for the copyright to expire and the book to enter the public domain. A public domain book is one that was never subject 

to copyright or whose legal copyright term has expired. Whether a book is in the public domain may vary country to country. Public domain books 

are our gateways to the past, representing a wealth of history, culture and knowledge that's often difficult to discover. 

Marks, notations and other maiginalia present in the original volume will appear in this file - a reminder of this book's long journey from the 

publisher to a library and finally to you. 

Usage guidelines 

Google is proud to partner with libraries to digitize public domain materials and make them widely accessible. Public domain books belong to the 
public and we are merely their custodians. Nevertheless, this work is expensive, so in order to keep providing tliis resource, we liave taken steps to 
prevent abuse by commercial parties, including placing technical restrictions on automated querying. 
We also ask that you: 

+ Make non-commercial use of the files We designed Google Book Search for use by individuals, and we request that you use these files for 
personal, non-commercial purposes. 

+ Refrain fivm automated querying Do not send automated queries of any sort to Google's system: If you are conducting research on machine 
translation, optical character recognition or other areas where access to a large amount of text is helpful, please contact us. We encourage the 
use of public domain materials for these purposes and may be able to help. 

+ Maintain attributionTht GoogXt "watermark" you see on each file is essential for in forming people about this project and helping them find 
additional materials through Google Book Search. Please do not remove it. 

+ Keep it legal Whatever your use, remember that you are responsible for ensuring that what you are doing is legal. Do not assume that just 
because we believe a book is in the public domain for users in the United States, that the work is also in the public domain for users in other 
countries. Whether a book is still in copyright varies from country to country, and we can't offer guidance on whether any specific use of 
any specific book is allowed. Please do not assume that a book's appearance in Google Book Search means it can be used in any manner 
anywhere in the world. Copyright infringement liabili^ can be quite severe. 

About Google Book Search 

Google's mission is to organize the world's information and to make it universally accessible and useful. Google Book Search helps readers 
discover the world's books while helping authors and publishers reach new audiences. You can search through the full text of this book on the web 

at |http: //books .google .com/I 






^^^RAB-Vt"" 



f \ 










HISTORY OF ENGLISH SOUNDS 



8WEET 



HENRV FROWDE 



Oxford University Press W, 
Amen Corner, E.C. 



A 



HISTORY OP ENGLISH SOUNDS 



FROM THE EARLIEST PERIOD 



WITH 



FULL WORB^ LISTS 



BY 



HENRY SWEET, M.A. 

BALLIOL COLUBOI, OXFORD ; HON. PH.D. HEIDBLBEKQ 



Oxfot6 

AT THE CLARENDON PRESS 

1888 

[ All rifjhts reserved ] 



c 
r 



1 



II 



^. ^. 



r^ - 



• ; 



;/ • 



P PEEFACE. 

y M 

I 

This work first appeared in the Transactions of the 
v.) Philological Society for 1873-4. Additional copies were 

struck off for the members of the Dialect Society. I also 
put a few hundred copies into the hands of Mr. Tnibner, so 
as to make the work accessible to the general public ; these 
have long since been sold out. 

My investigations were due to the combined influence of 
Bell s Visible Speech, Ellis's Early English Pronunciation, 
and the German school of comparative and historical philo- 
logj) of which Orimm*s BeuUche Orammatik was to me still 
the chief exponent. In attempting to trace the connection 
between the Old English vowel-system as revealed by its 
complirison with the cognate languages, with the early 
Modem one deduced by Mr. Ellis from his contemporary 
phonetic authorities, certain difficulties became manifest. 
It was evident that if the present distinction between 00 
and (7, as in moan^ stone, already existed in the Old English 
monay sfdn, it must also have existed in the intermediate 
Chaucer period, so that Chaucer's in mone, ston could not 
possibly have stood for one and the same sound, as Mr. Ellis 
assumed. This self-evident objection to Mr. £llis*s view had 
also been made by Dr. Weymouth, and had been, indeed, 
foreshadowed by Rapp (see Ellis EEP 11, 675). But with 
the es the parallelism did not seem to hold good. The cor- 
respondence was clear enough in the case of Old English 
<?, eo, ea, but Old English ^ seemed to be represented by ee 
and e, ea at random, so as partly to justify Rapp's and Ellis's 






VI PREFACE. 

levelling of Chaucer's long e& under one sound. On 
further examination it turned out, however, that Old English 
^ became Modem English ee only in those words in which it 
corresponded to Gothic e, as in deed. The inference was 
clear, viz. that in the dialect which gave rise to Standard 
English the West-Saxon a in dad must have been represented 
by e. The difficulty of the lengthening of the close Old 
English e of stelan into Modem English ea was cleared up by 
a comparison with similar changes in the Modem Scandi- 
navian languages. Rough and tentative as these investiga- 
tions were, they sufficed to show that the development of 
English sounds followed definite laws, and was not — ^as had 
hitherto been tacitly assumed — the result of mere chance and 
caprice. 

My History of English Sounds was, therefore, originally 
an expansion of a monograph on the history of long e and o. 
At first intended to treat only of the vowels, it was afterwards 
made to include a sketch of the consonants. The word- 
list grew under my hands in the same way. 

The defects of the book were the result of the inevitable 
gaps in the knowledge of an entirely self-trained student. In 
1873 my undergraduate career at Oxford had only just come 
to an end, and Plato and Aristotle had so interfered with my 
own proper studies that my knowledge of Old English was 
at a lower ebb than it had ever been during the preceding 
five years. My ms dictionary was stiU only in the index 
stage, and the considerable stores of material it contained 
were therefore practically out of my reach, and I was obliged 
to rely mainly on Ettmiiller's lexicon and my own memory. 
Middle English I had hardly studied at all. 

As might be expected, I had failed to keep pace with the 
advance of German investigation. I still kept the antiquated 
view of the priority of Gothic i in stilan etc. I had glanced 
through Scherer's Zur Geschickte der deutschen Sprache^ but my 
dislike to his theory of * tone-raising ' — well-founded as it 



PREFACE. Vll 

was — unfortunately prevented me from appreciating his 
explanation of the development of Old English ea. 

The most serious of my defects of method was my rejection 
of the principle of gradual sound-change in favour of change 
j>er sallum (EUis EEF i8), although in practice (as in treating 
of the diphthonging of i) I admitted that these leaps were 
infinitesimally small. Of the influence of stress in forming 
doublets etc I had no idea, although a phonetic student of 
living English might well have partially anticipated the 
later investigations of Osthofl*, Paul pnd their fellow-workers. 

I had, on the other hand, clearly grasped the distinction 
between phonetic and analogical sound-change. It seems 
now self-evident that the preterite bore owes its o to the past 
participle ; but when I first propounded this view before the 
Philological Society, it met with opposition : people ' didn't 
see why ' Old English a should not become o merely because 
the Germans said it ought to become a. 

Things have changed in the last fifteen years. The adop- 
tion of German methods is no longer a bar to recognition 
and success. Now too that the Germans are beginning to 
take up practical phonetics, its importance is beginning to 
be recognized in the land of its birth. German philology 
itself has been quickened into new life. English philology 
has been made a specialty in the German universities : it 
boasts a ^literature' of its own; it is even beginning to 
develope cliques and schools. 

Nor have I been idle myself. My Old English dictionary 
collections have been brought into more manageable shape, 
and have been supplemented by similar collections from the 
Middle English texts. My range of languages— both dead 
and living — ^has been widely extended, with a corresponding 
advance in my command of sounds. I have done my best 
to keep level with the latest results of foreign investigation. 

Hence the present second edition, while adhering to the 
general plan of the first, is in execution and detail an entirely 



VUl PREFACE. 

new work, which has not only been re-written from beginning 
to end, but is based on a fresh collection of materials. 
Its object is, however, the same as before — ^to sketch the 
development and history of English sounds fi'om the very 
beginnings of articulate speech down to the present day, 
with such discussion of the general principles of sound- 
formation, sound-change, sound-representation and the de- 
velopment of dialects and languages as seemed necessary. 

It is evident that so ambitious and comprehensive a scheme 
as this can only be carried out by subordination of details to 
general principles, and strict adherence to the main line of 
development. This main line of development itself need 
not and cannot be traced with equal fulness throughout. 
A history of English sounds which did not go further back 
than Old English would still be complete in itself, and might 
well content itself with a reference to other books for the 
Germanic and Aiian sounds, which cannot be adequately 
treated of without going into the details of a considerable 
number of separate languages. Brugman's Grundris^y which 
is confined to the phonology of the Arian languages, takes 
up nearly six hundred pages, and yet it is — what it professes 
to be — a mere outline 1 But the main features of Arian 
phonology can be stated in a much smaller space, if the 
reader is contented to take them without detailed proofs. Such 
brief summaries of Arian and Qermanic phonology as I have 
given in this book are, besides, useful for reference even to 
those who have studied special treatises on these subjects. 
At the time, too, when I wrote this book, Brugman's Grundriss 
had not appeared, and a knowledge of the latest results had 
to be laboriously gleaned from a variety of sources — often 
almost inaccessible to an Englishman. The ten pages 
into which I have condensed my sketch of the Arian sounds 
represent years of tedious toil and groping after light. 

The most serious defect of the book is that I have not 
been able to make any general use of the modem English 



PEEPACE. IX 

and Scotch dialects, which (with a few brilliant exceptions) 
have been treated in such a way as to make them worse than 
useless for purposes of historical phonology. American 
English and Irish English are equaUy important and equally 
inaccessible at present. 

In the present edition I have made less use than in the 
first of the living Germanic languages. The reason is that 
I feel too painfully the defects of my knowledge of them. 
In 1877, when my practical knowledge of them was still 
fresh, I wrote out for press a sketch of the comparative 
grammar of the six Hterary Germanic languages in their 
spoken form, but was unable to find a publisher, and the 
work is, of course, now antiquated. 

I have abstained throughout from controversy or discussion 
of doubtful points, as far as possible. I have tried, to the 
best of my ability, to arrive at an independent judgment on 
each question by an impartial study of the evidence and 
the views of my predecessors and contemporaries. I have 
also abstained, as a rule, from giving references to the works 
of others, or attempting to settle questions of priority of 
discovery : this I leave to the future historian of nineteenth 
century philology. I will only add that many of the new 
views expressed in this work were first published (generally 
in a very brief form) in the proceedings of the Philological 
Society, where also may be found Henry Nicol's valuable 
contributions to the history of Middle English sounds and 
orthography. 

The reader will observe several novelties in terminology, 
especially in the section on Sound-Change. I use ^Arian' 
instead of the clumsy * Indo-Germanic ' ; as the word aria 
is alway three-syllabled in Vedic Sanskrit, I see no reason 
for writing it * Aryan.' 

My use of the revised Visible Speech notation for exact 
purposes requires no justification. Although far from perfect, 
it is the only system which is universal in its application 



X PEEFACB, 

and at the same time capable of being worked practically. 
Although experience shows that there is no chance of philo- 
logists agreeing on a general Roman system, I have given one 
in the chapter on Phonetics for the use of those who have 
not access to the Visible Speech letters. It will be observed 
that I use the less accurate ^ Broad Bomic' as a kind of 
algebraic notation, each letter representing a group of 
similar sounds. 

In the chapter on Sound-Change I have aimed more at 
reliability than fulness : nearly the whole of the material is 
drawn from languages of which I have a practical know- 
ledge. 

In Sound-Representation the section on the Laws of Form- 
Change was suggested by the observations I made in working 
out a system of Shorthand on which I have been engaged for 
some years. That on Alphabets is based partly on an elaborate 
study of Old English palaeography which I made many years 
ago, partly on Wattenbach's Latdniscke PalcRographie, 

In treating of the Runes I have followed Dr. Wimmer's 
Runeskriftens Oprindehe (now accessible in a German transla- 
tion) very closely. But at the same time I could not help 
feeling the force of Canon Taylor's arguments against the 
Latin origin of the runes, as stated in his Greeks and Goths. 
So I have had to steer a middle course in this hopelessly 
obscure question. 

In the Modem English section I have relied for my material 
almost entirely on Mr. Ellis's Early English Pronunciation, 
To save the reader the trouble of constant reference to this 
great work, I have given the statements of Mr. Ellis's 
phonetic authorities in full, wherever necessary. 

Especial care has been expended on the First Word-list, 
which is based mainly on my own collections, as far as Old 
and Middle English are concerned. For the early Modem 
period I have, of course, relied mainly on Mr. Ellis's pronounc- 
ing vocabularies. The Middle English quotations make no 



PREFACE. Xi 

pretension to completeness. At first it was my intention to 
confine myself to three or four representative texts, but, as 
might be expected, I found it advisable to widen my range 
as I went on. An ideaUy perfect list would, of course, give 
the forms not only of all the dialects of the three periods of 
English, but also of the cognate Germanic and Arian lan- 
guages, together with references to the body of the work for 
explanations of obscure or abnormal developments, and would 
also include proper names, which I have only occasionally 
dealt with. 

It is evident that any attempt to carry out this ideal would 
have involved another ten or fifteen years delay in bringing 
the book out, and would have swelled its bulk to an inde- 
finite extent. In the present unsettled and progressive state 
of English philology and the utter uncertainty of its prospects 
in this country, excessive elaboration would be a waste of 
time. I am as fully alive to the defects of this work as any 
of my critics can be, but nevertheless when I see the great 
advance it is on the first edition, I cannot help regarding it 
with a feeling of satisfaction, which is not diminished by the 
reflection that its best portions are, after all, little more than 
summaries of the work of others. It is to me a source of some 
pride that, just as Henry Nicol and myself were the first to 
take up Bell's Visible Speech and apply it to linguistic investi- 
gation and the practical study of language, so also we were the 
first to welcome the revolutionary investigations in Ellis's Early 
English Pronunciation, and were the first in England (with the 
brilliant exception of John Kemble) to apply German methods 
to English philology, although from the beginning we set our 
faces against the * woodenness ' which then characterized Ger- 
man philology : its contempt for phonetics and living speech 
and reverence for the dead letter, its one-sidedly historical spirit, 
and disregard of sJiSkLogy. So too at a later date, I was one 
of the first in England to welcome the ' neo-philological ' 
reformers who have rescued German philology from its earlier 



Xll PBEFACE, 

stagnation of methods. Of the many illustrious members 
of this school I owe most to Paul and Sievers. No one can 
read the chapters on general principles in my book without 
seeing how much I owe to Paul's Principien der Sprachgeschichte, 
My debt to Sievers^s Phonetik is seen in the chapter on Sound- 
Change, and in almost every paragraph that deals with Old 
English ; Sievers's Angehdchsisehe Grammatik has indeed lighted 
up the obscure and tortuous paths of Old English dialectology 
and linguistic chronology in much the same way as Bopp's 
grammar lighted up the intricacies of Arian philology. I 
only regret that by an unfortunate accident I was prevented 
from utilizing the second edition of Sievers's grammar ^. My 
debt to Mr. Bell speaks for itself. My debt to Mr. Ellis is best 
expressed by repeating what I said in the Concluding Remarks 
to my first edition : " As regards my obligations to Mr. Ellis, 
I can only say, once for all, that without his investigations this 
essay would never have been written. It is essentially based 
on his results, of which, in some places, it is little more than 
a summary ; while I have throughout drawn largely on the 
enormous mass of material stored up in the ' Early English 
Pronunciation'.'' If I had to dedicate this book, it would 
receive on its title-page the four names of Bell, Ellis, 
Paul, and Sievers. 

HENRY SWEET. 

Bath : lUh January ^ 1888. 

' I mAy take this opportunifcy of saying tbat I have definitely abandoned my 
intention of bringing out a grammar to my Oldeit English Texts, Those German 
scholars who have hitherto refrained from utilizing tbat work for grammatical 
investigations need no longer have any scruples on my account. 



CONTENTS. 



-♦♦- 



Phokxticb page 

Analyiis i 

Synthesis ^ 

Notation 13 

Soukd-Chanob 14 

Internal leoUtiTa 

Breath and Yoioe 18 

Vowels 19 

Ck>nsonants 23 

Quantitj 29 

Foioe 30 

- Intonation Z^^*^^ 

Transposition 33 

Internal Combinative 

Breath and Voice 34 

Front-Modification 36 

Back Inflaenoe 38 

Bounding 38 

Nasalizing 38 

Parasiting 40 

Other Inflaenoes 41 

Aooustio Changes 43 

Bztemal Changes 45 

Qeneral Frlnoiplea 49 

Oriqin op Speioh-Sounds 50 

OBIOlir OP DiALICTS 52 

Sound Rbpbisbntation 

Origin of Writing 59 

Laws of Form-Change 60 

Alphabets 63 

New Letters 65 

Correspondence of Sound and Symbol 67 

Normalising 70 

Synthesis 71 

Interpretation of Symbols 72 



XIV CONTENTS. 

Abian Sounds 74 

ConflonAXLto 83 

Gkbxanio Sounds 85 

VoweU 86 

ConBonftntfl 87 

High-German Consonant-Shift 93 

Runes 

Germanic 95 

Old-English 97 

Old-English Sounds 

Dialects and Texts 99 

Orthography loi 

Stress (Metre) 102 

Quantity 106 

Vowels 116 

Consonants 134 

SCANDINAYIAN SOUNDS 

Orthog^phy 150 

Vowels 150 

Consonants 151 

Inflaenoe on English 153 

Middle English Sounds 

Dialects and Texts 154 

Orthog^phy 156 

Metre and Stress 163 

Quantity 165 

Vowels 171 

Consonants 189 

MoDEBN English Sounds 

Periods 199 

Phonetic Authorities 202 

Orthography 208 

Short Vowels 209 

Long Vowels 239 

Diphthongs 241 

Consonants 359 

Living English Sounds 272 

Stress 273 

Quantity 274 

Vowels 275 

Consonants 277 

First Woed-List 279 

Second Wobd-List 373 

Index to First Word-List 394 



CONTENTS. XV 

PAGB 

Tablbs 

I. Sonnd-ohange 402 

II. Fomu of Letters 404 

III. English Vowels 405 

IV. Old-English Dialects ' .... 406 

V. Middle-English Dialects 406 

VI. Modem English Vowels 407 

CoNTBACTioira 408 




ERRATA. 

§ 32. 8. tran$poie riBing and falling. 

41. 7. for that read thoae. 

168. a. oifit^ Greek j)^to. 
p. 48. 10. for dasennat read dosexmat 
§ 186. 3. for (ij) read (g). 

302. 34. for inner read outer. 

306. 13. /or {o)a read o(a). 

384. I. for jimim read jirni(m. 

353. 10. for f reod I. 
Sis, 6. for E. read E, ^* 



HISTORY OF EKGLISI SOUNDS. 



-•4- 



PHONETICS ^ 

ANALYSIS. 

1. In the following sketch the revised Visible-Speech 
symbob ' are employed for exact notation, with an occasional 
Romic transcription. The Romic symbols are enclosed in ( ) 
where necessary to prevent confusion. They are sometimes 
used more loosely, especially in representing the sounds of 
dead languages (37). 

2. Speech-sounds are generally formed with out-breathing 
or expiration (>), rarely with in-breathing or inspiration (<)• 
Suctum-dop^ or clicks, as in the f&miliar tut ! D«, ai*e formed 
without either out- or in-breathing. 

8. Throat SoundB. When the glottis is wide open, the air 
passing through it produces breath (o); when the glottis is 
narrowed so as to make the vocal chords vibrate, voice (i) is 
the result; if the chords are approximated without being 
allowed to vibrate, whimper («) is the result. If whisper is 
strengthened by contraction of the superglottal passage or 
'false glottis,' we get the wheeze («a), as in the Arabic Hha^ 
which can be voiced (♦), as in the Arabic Ain. The Glottal 
etop (x) is produced by a sudden shufcting or opening of the 
glottis, as in a cough. 

4. Nasal Sounds are formed by depressing the uvula so as to 
let the breath pass through the nose. Nasality is denoted by I. 

5. Narrow and Wide. Narrow (:) sounds are formed with 

' Sweet : Handbook of Phonetics. Sieyers : Fhonetik. Vietor : Elementa der 
Phonetik. 
' See my Bound-Notatton in Tnuue. of PhiL Soo. 1880-1, JI. 



2 HISTOET OP ENGLISH SOUNDS. 

tensity and convexity, wide (x) with slackness and flatness of 
the tongue. There are various degrees of narrowness, and it 
is possible to produce a sound which is exactly half way ; the 
Norwegian short i in fisk is an example (I:). 

6. Vowels are voice modified by different configurations of 
the superglottal passages, but without audible friction. By 
position they are either back (guttural), /ro»^ (palatal), or mixed,, 
that is, formed by a position intermediate between back and 
front. They have three degrees of elevation of the tongue, 
high, mid, low. When the tongue is lowered from the high 
position, the place of narrowing is at the same time shifted 
further back. So we have altogether nine positions : 

high-back high-mixed high-front 

mid-back mid-mixed mid-front 

low-back low-mixed low-front 

Each of these positions yields a different vowel-sound accord- 
ing as the tongue has the narrow or the wide shape. Inter- 
mediate positions are : retracted (4) and advanced (►), raided ('^) 
and lowered (t). Each vowel-position can be further modified 
by rounding (labialization). Front vowels are rounded by the 
lips only (outer rounding), mixed and back vowels more by 
the cheeks (inner rounding). There are three degrees of lip- 
and cheek-contraction in rounding, high vowels having the 
narrowest, low vowels the widest lip-opening. When a 
vowel has a higher degree of rounding than belongs to its 
height, aa when a mid vowel is formed with the rounding 
of a high vowel, it is said to be over-rounded, which is denoted 
by adding the 'rounder,* as in })=the Swedish close 0. The 
opposite phenomenon of under-rounding is denoted by adding 
the 'rounder' to the symbol of a front vowel, the 'inner 
rounder' (^) to that of a mixed or back vowel, as in 1)=: 
Swedish y, 1»= Swedish short u. Vowels are also capable 
of jt7^»^-modification (v), the tip of the tongue being raised 
while the vowel-position is maintained. 

7. The thirty-six elementary vowels are given in the an- 
nexed table in their Organic and Romic symbols, together 
with key- words : 



^2'|.# 3 14 



it 


k 


t S" 


iiji 




^11 



4 HISTOEy OF ENGLISH SOUNDS. 

8. Acoustically considered, a vowel is voice modified by a 
resonance-chamber, viz. the mouth. Every time we alter the 
position of the tongue and lips, we create a new resonance-* 
chamber, which moulds the voice into a different vowel. 
Each vowel has an inherent pitch, due to the shape and 
size of the resonance-chamber. The following is the order 
of some of the chief vowels in pitch, beginning with the 
lowest: 

*; *; }; J; |; i; j; ]; 3; i. tvLfj i>*; x> v C; [; J; f 

u; t^; o; O', o; o; v; &; a; a,(b; e,Q; i,j; aisd; e; e; i; i. 

Vowels separated only by a comma have the same pitch. 
It will be observed that vowels of distinct formation are 
often alike in pitch and consequently in sound (174). 

9. Consonants are the result of audible friction or stopping 
of the breath in the throat or mouth. But in many con- 
sonants the friction is not audible when they are uttered 
with voice. When the friction is audible in a voiced con- 
sonant, as in 8 2r, > 9, it is called a buzz^ the corresponding 
breath consonants s s, >/, being called Aisses, All consonants 
can be formed either with breathy voice^ or whisper. The last 
are denoted by i: :>»= whispered /. Consonants are either 
narrow (j) or wide (:) ; in E. they are wide, «; w being equiva- 
lent to dose (10) 1 {u). 

10. By form there are four classes of consonants: (i) opm^ 
such as (Dr, 6«. {%) side^ such as oo /, which is often one-sided 
or unilateral. (3) stopped^ such as Qi^md. (4) nasal, formed 
with stoppage of the mouth passage, the nose passage being 
left open, as in m^rm. When an unstopped (open or side) 
consonant is pronounced with the nose passage open, it is 
said to be 'nasalized,' as in 85, which is a nasalized 10 or an 
m uttered with only partial lip-dosure. Trills or 'rolled' 
consonants are a special variety of open consonants, and are 
denoted by 1 ; thus m is the Scotch r, AU consonants may 
be pronounced with tenseriess (a) or 'closeness,' or with looseness 
(v), thus a loose y ov is equivalent to the vowel I (i). 

11. By place there are five main classes, (i) back, such as 
a i, d 1} (as in sing). {%) froiU, such as O)/ (in you), (3) poitU, 



PHONETICS. 5 

such as Dt, to I. (4) blade (formed by the surface of the 
tongue immediately behind the point), such as S«. (5) lip^ 
such as op, rtn. Point and blade consonants are included 
under the general term ^ foreward.' Most of these admit also 
of ' inner' and ' outer ' varieties, as in the case of the vowels. 
Point consonants admit of inversion (c), in which the point of 
the tongue is turned back, and jMrotrtmon (3), in which it is 
protruded to the lips. Thus Oc is an inverted or 'cerebral' t. 
Some consonants are formed by a combination of two posi- 
tions. Thus ej (as in she) is a blade-point consonant, >/ a 
lip-teeth consonant, v/ (as in thin) a point-teeth consonant, 
which is really equivalent to * outer point.' When the point 
of the tongue is put between the teeth, the sound is called 
' interdentaL' All consonants are liable to be modified by 
the back-open ( (a?), front-open \ (y), point-open v (r), lip-open 
)(t;)=' outer rounding,' and lip-back-open p(«^)=* inner round- 
ing' positions. C)=G, ch in at^h is for convenience written 
c, and 3(=E. toh is written d. Other combinations are ex- 
pressed by * between the symbols of the two elements, as in 
0*0= h and p uttered simultaneously, or by means of the 
blade- (i), stop- (1), open (11), side- (h), tmilateral («), throat-stop- (1) 
modifiers. {*) is used as a general modifier, thus u^ is any 
variety of /. 

12. A general table of the consonants is given on the fol- 
lowing page : — 



HISTOEY OF ENGLISH SOUNDS. 





•^ 






















^fca 






















^ 






















^ 






















1 

• 


C4H 










> 










S 


A 










A 








^ 












1 








s 






















^ 


^ ii 












1 








• 


44 










^ 


1 

1 








>^ 


Q » 










Q 


1 


1 














> 










it 


•©• 




Q^ 


1 








^ 


S 









C9 





I. 




Ocg 


00 





IL 


•^ 




















T* 












§. 










^ 


« _.**^ 










K3 9 










Bq 


«>ca 




•5 






« S 




s 




1 


















g 


fi^ 


(0 








• 

M 



to 




B 


1 


1 

1 


















s 




d 


fit*; 


r 






d 


B 


r 




»s 






fl 












•5 


A 






^ 














Bi 


"S*^ 

^ S 


3 


4a 


"^1 




b 


^ 


^ 


a 


• 




o»^ 


a 


D 


r*^ 




3 


8 


B 


r 








>^ 








J^ 






*«« 






■& 








f»M 


bo 


a 




§ 


•^ 




m 








&i) 


• 


60 




fS; 


-•^. 


a 


0^ 


;§ 




•-^g 


8^ 


8^ 








c^ 


c 


afti 


«j 




e§, 


q!^ 


o-^ 














08 




















d 


sS 








1 


1 

H . 


% 


M 


•^ 




1 




t« 


c» §^ 






0^ 


CO 





-» 




ttO 


•(i 





^ 00 


1 


08 


• 
• 
• 




• 
• 
• 


'3 


• 
• 
• 


• 
• 


• 
• 
• 












•>* 








»»<* 






1 


^ 


eg 


1 






•§ 

^ 


1.^ 
« 

^ 


1 










t 















PHONETICS. 



Id. Each consonant has an inherent pitch of its own. The 
following are the pitches of the chief open bi'eaih consonants, 
beginning with the lowest : 



o 


a 


c 


> 





o 


e 


s 


V 


o 


wA 


xw 


X 


f 


* 


ri 


J 


8 


y 


« 



14. There is a close relation between consonants and 
vowels. In many open Yoiced consonants there is no audible 
friction, and such 'vowellike' or 'liquid' consonants have 
quite the effect of vowels. These are C"^, <f>j\ ow; ior,(Dl and 
the nasals. But e and o) can also' be buzzed. The two closest 
vowels I (i) and 1 (u) approximate so closely to the consonants 
(D and o respectively, tiiat it is often ^fficult to draw the 
line. When devocalized these vowels cannot be separated 
from o and d. The following are the most important of the 
relations between individual consonants and vowels : 

e-i € ei> (D-i 0) 0)*' 6-I e: 6^, o (d)-i (D) o))** 
J ] 1 I C ^ J } i t { f 

(D r may also be weakened into a kind of vowel ; in fact, the 
E. r in very may be considered as an unsyllabic vowel. 

15. The acoustic relations between consonants and vowels 
may be seen by comparing the tables of pitches. They 
generally agree with the organic relations. Observe that 
8 and i are acoustically similar. 



SYNTHESIS. 

16. Quantity. For general purpolses it is enough to 
distinguish three degrees of quantity or length: lon^ («), 
half-long (*), and %h(yrt (t), the last being generally left un- 
marked. In practice the distinction of long and short is 
generally enough. Long vowels are doubled in the Romic 
notation. 

17. Force. Loudness and stress (accent) depend on the 
force with which the breath is expelled (generally from the 



8 HISTORY OF ENGLISH SOUNDS. 

lungs). In a single breath-impnlse, as in the vowel aa, we 
can have three kinds of force : 

level ]♦=: 

increasing 3*< 
diminishing ]«> 

The tendency in language is to utter with diminishing 
force. 

18. The influence of force on the synthesis of speech is 
very important, for the sense of unity and separation depends 
mainly on it. Continuity of force gives a sense of unity ^ as in 
]♦>, ^O^ discontinuity, as in ^», that of separation, the ]4 
being broken up into two syllables. Hence, every syllable 
(vowel-group) must be uttered with a single impulse of 
breath, as it would otherwise be split up into two. In 
language the tendency is against uttering two successive 
syllables with the same force. 

19. The comparative foroe with which the separate syllables 
of a sound-group (word, clause, or sentence) are uttered is 
called stress (accent). There are three main degrees of stress : 
strong (•), half-strong or 'medium' (:), and weak (*), the last 
being generally left unmarked. Weak stress is also marked 
by (-)• In practice it is often sufficient to mark the strong 
stress only. The stress-marks are put before the element on 
which the stress begins. The tendency in language is to 
alternate strong (or medium) and weak stress. Thus, if such 
a group as kalana is stressed on the first syllable, the second 
is generally weak, the third medium or, at any rate, slightly 
stronger than the second: *q3*od3:73- But in rapid speech 
such a word might also be pronounced 'a]*CD3^3> ^^ & single 
impulse of breath. The answer to the question, Where does 
the syllable begin ? is, that if it has a distinct stress (strong 
or medium) its beginning corresponds with the beginning of 
the stress. If, on the other hand, the syllable is weak, it is 
often difficult to settle where it begins. Hence it is possible 
to alter the syllable division by shifting the stress from one 
element to another. Thus at all ought strictly to be pro- 



PHONETICS. 9 

nounced lO'jtoo, but in actual speech the second syllable 
begins on the t: V^^* 

20. The distinction between long and double consonants 
also depends on stress and syllabification: in '3cd*:3 the 
consonant is long, in *3qo:Oo3 or 'S^")^] ^^ ^^ double. Double 
consonants cannot occur finally or isolated. 

21. A sound which can form a syllable by itself is called 
9yllabic, Syllabicness implies an appreciable duration and 
force. The distinction between syllabic and non-syllabic is 
generally parallel to that between vowel and consonant. 
But those 'vowellike* or 'liquid* voiced consonants which 
are unaccompanied by buzz are often also syllabic. These are 
a)r, 00^ and the nasals. Even voiceless consonants can be 
syllabic, as in ds«o pst^ where the s is syllabically equivalent 
to a vowel by virtue of its length and stress, the unsyllabio 
D and o being comparatively momentary and stressless. A 
syUabic consonant is denoted by ] : 0x000]= 'battle.' 

22. A vowel, on the other hand, can lose its syllabicness, 
especially in combination with another vowel, with which it 
then forms a diphthong. These diphthongic or 'glide-' vowels 
are written consonant size in YS, being, from a syllabic point 
of view, consonantal vowels, as in 3^ at, where the group is 
uttered with one impulse of diminishing force, ij ia, which 
implies increasing force, the latter diphthong being almost 
equivalent to 0)3 jet* 3^ is called a ' rising,' s!\ a ' falling ' 
diphthong. ^ with the second element lengthened ought to 
be considered a dissyllable, but it has the efiect of a diphthong 
if the Xk is kept stressless. 

28. Glides. In pronouncing any sound-combination, such 
as al ^1, we not only have the two sounds Q and I, but also 
the ' glide,' or sound produced in passing from the one position 
to the other, which is implied by the juxtaposition of the 
symbols. The glide is called the ' off-glide * of a, the * on- 
gHde ' of f . If the transition is made slowly, the glide be- 
comes so distinct that it becomes necessary to write it 
separately. Thus 3^3 V^ ^^7 ^ developed into full 3x(i>3} 
3x0)3 ^if^' OUdeness and nonsyllabicness generally go to- 
gether, but it is often difficult to draw the line between 



lO HISTORY OF ENGLISH SOUNDS. 

gliding and fixed configuration, especially in the consonants. 
Gliding quality is marked ), when necessary, as in oo)(d'|;Ct= 
' try,' where o and 0) together have the same length as the 
single (I) in aKDl^T=*dry/ 

24. Initial and Final Fowel-glides, Vowels may be begun 
and finished in various ways : 

(a) The glottis is gradually narrowed, passing through the 
various positions for breath and whisper till voice is pro- 
duced. This gives the ' gradual ' beginning -O*. 

{b) Tlie breath is kept back till the glottis is closed for 
voice, which begins at once without any introductory breath. 
This is the * clear ' beginning 'J*. 

25. In both these cases the stress begins on the vowel. 
If it is thrown on to the preceding glides, they are at once 
recognized as independant elements, *3^ becoming o^ haa^ 
with the ordinary * aspirate,' while '3* becomes xjj*, with the 
glottal stop, o is generally modified by the following vowel, 
whose configuration it partly anticipates. It is, therefore, 
the voiceless glide vowel-vowel corresponding to the vowel 
which follows, and is then written fi. fil, si hi^ hu are, in 
fact, equivalent to x:I, i:l, being almost equivalent to weaker 
forms of ol, oi jhi, whu. 

26. Vowels are finished analogously : 

(a) by a gradual opening of the glottis, the final glide 
passing through whisper to breath, giving the 'gradual' 
ending I*. 

{b) by a cessation of expiration while the glottis is still 
closed for voice, giving the * clear' ending I'. If uttered with 
stress these endings become respectively i^ or Is and Ix. 

27. Consonant-glides. All consonants consist of three 
elements, (i) the consonant itself, (2) the on-glide, and 
(3) the oflT-glide. Each of these elements may be either 
breathed or voiced, and may be modified in various ways. 
The off-glides of stops are the most important. 

28. The following are the combinations, as regards breath 
and voice : 





PHONETICS. 




I5ITIAL. 


MSDIAL. 


FIITAL 


a-3 


3a-3 


3a- 


a'3 


3a'3 


3a' 


a-3 


3a-3 


3a- 


a'3 


3a'3 


3a' 



II 



O* is the E. k in all positions, G* the £. final y, as in egg^ and 
Q* the E. medial g^ as in eager. Q' is the Middle and South 
German k. E. initial g^ as in go^ is often nearly Q', but there 
seems to be a trace of vocality in the stop itself. On-glides 
after vowels are generally voiced, but are breathed in some 
languages, as in Icelandic : J'Q*, ]*a*. 

29. All stops, especially when voiceless, postulate a certain 
compression of the breath behind the stop, so as to produce 
an audible explosion when the stop is removed. However 
strong this explosion of a breath-glide, it is not felt as an 
independent element, unless the initial force is maintained 
during the formation of the glide itself. In this latter way 
are formed the Danish aspirates Qo, etc., as in komme, 

80. The glides of unstopped consonants are less marked, 
but the vocality of the consonants themselves is, on the 
other hand, more distinct than in the case of the stops. 
8 2r, etc., admit of 'gradual' and 'clear' beginnings and end- 
ings, analogous to those of the vowels; final z in E. has 
the gradual ending -86)*. After another buzz or voiced stop 
it is completely whispered in E., as in fi|Ja)S> ' heads.' 

8L Consonant-glides may be variously modified by round- 
ing, etc. Thus E. ' cool ' is really a*»i900. We can also have 
such a combination as Q*»], distinct from ao] (k^^), which 
is equivalent to Qo»3* In such cases the rounding is generally 
begun before the stop is loosened. 

82. Glideless CombinationB. In speech the general prin- 
ciple is to take the shortest way between two sounds. This 
often results in combinations which are effected without any 
glide at all. This is regularly the case in sequences of 
consonants having the same place and differing only in form. 
Thus in passing from i » to o ^ in 7(D all that is done is to 
close the nasal passage. Similarly, in (DQO dl the transition is 



12 HISTORY OP ENGLISH SOUNDS. 

made by simply opening the side apertures, the tip of the 
tongue retaining its position. Combinations in which a stop 
is followed by open consonants formed in the same, or nearly 
the same, place are effected either with no glide at all, as in 
DO jD<^, or with a very slight one, ba in o> pf. In such com- 
binations as ti, tf the places of the two consonants are gener- 
ally approximated as much as possible, so as to get rid of 
the glide, thus E. t% is really OfS or sis, E. eh o^^. Even when 
consonants formed in different places come together, it is 
possible to combine them without any glide, although in 
these cases the gliding must be regarded as the normal form. 
Absence of glide is marked (•). Thus E. <ict is X^'^i ^be tip of 
the tongue being brought into position before the a-contact 
is loosened, while in French and other languages there would 
be a breath-glide between the two consonants. 

38. Glide-oonsonants in the special sense of the word are 
consonants formed without any fixed configuration, however 
much the transition may be prolonged. The most distinct 
glide-consonant§i are the^jt^, of which the Norwegian Hhick' 
/ is an example : an inverted r finished off with a momentary 
contact of the tongue-tip with the inside of the palatal arch, 
the tongue moving forward all the time. 

84. Intonation. Changes of pitch or tone may proceed 
either by leaps or glides. There are three primary intona- 
tions : (i) the level (— ), (2) the riiing ('), and (3) iiie Jallinff f ). 
There are also compound tones, formed by uniting a rise and 
fall in one syllable: (i) the ccmpound-'rmnff (*), (2) the com- 
pound-falling (*). 

85. A level tone can be of any height^ but it is enough to 
distinguish high-level (") and low-level (.). The gliding 
tones can also begin at any height — ^low-rising Q, high-rising 
('), etc. They can also be varied indefinitely according to the 
interval through which they pass. 

86. Besides the separate intonations of which it is com- 
posed, each sentence, or sentence-group, has a general pitch 
or key of its own, which may be high or low. Changes of 
key may proceed either by leaps or by glides. 



PHONETICS. 1 3 



NOTATION. 



37. The Visible Speech symbols have been fully explained 
in the preceding sketch. It is, however, oonvenient to have 
a more general notation, in which only the broader distinc- 
tions of sound are recognized. The following are the vowel- 
symbols of such a 'Broad Bomic' notation, which can, of 
course, be supplemented by the more exact symbols already 
given : 

a ai in father. 
£B „ „ man. 
e =zclo9e [or open) e. 
e zzzopen e. 

e :=ant/ mixed or obscure vowel. 
i 09 in it. 

o =cla8e (or open) o. 
0, o=open o. 

ce =eloee (or open) Germ, o. 
ce =open Germ* o. 
y =-Fr. u. 

Length is denoted by doubling. 

88. In dealing with dead languages, it is generally most 
oonvenient to give their spelling unaltered, except by the 
addition of diacritics, as in a=s(aa), ^=open e^ p=:open (?, 
«=sFr. tr, c, ^=0, ffl resp. 



14 HISTOEY OF ENGLISH SOUNDS. 



SOUND -CHANGE. 

80. Before entering on the subject of sound-change, it will 
be desirable to discuss the general question of word-division. 
The popular division of the elements of speech into sounds 
(^letters'), syllables, words, and sentences, is not purely 
phonetic, but also partly graphical and logical^ especially 
as regards word-division. No amount of study of the mere 
sounds of a sentence will enable us to recognize the words 
of which it consists. The only division actually made in 
speech is that into breath-group^^ due to the organic necessity 
of taking breath, which breath-groups correspond partially to 
the logical division into sentences. Within each breath-group 
there is no more pause than between the syllables of a single 
word. Thus, to the ear the word 'teller' and the sentence 
'tell her' are identical in sound — o|)a)X> cii^d we cannot 
possibly analyse such a sound-group without knowing its 
meaning, and even then word-division is a complex problem. 

40. At first, all sound-changes are carried out consistently 
through each breath-group, without regard to word-division. 
This primitive stage is clearly shown in the Celtic mutations. 
Thus in Welsh the change of jd into h between vowels is 
carried out not only in single words such as aber from Old 
Welsh aper 'confluence,' but also in such groups as dy ben 
* thy head ' = dy *pen. The result of this and similar influences 
is that the Welsh word for ' head ' appears in four different 
forms : pen DfT, ben dJJt, phen >[!, mhen [^[1, according to the 
original ending of the word preceding it. Now the logical 
side of language tends to rebel against such a multiplicity of 
forms, and in most languages we might predict that that form 
which is used at the beginning of a breath-group, viz. pen^ 
would gradually supplant the three others, dy ben, for in- 
stance, becoming *dy pen. In Welsh, however, these muta- 
tions were found useful for various grammatical distinctions — 
fawr >3^(D, for instance, being in certain collocations the 
feminine of mawr 'great' — and hence were preserved. 



SOUND-CHANGE. 1 5 

An equally -primitive stage is preserved in the Sanskrit 
sandhi, only here it is generally the end of a word that is 
modified, as when ^tdtas ca becomes tdtag ca oJoSo q3, the 
beginning of the next word being also modified in some eases, 
as when tdd grutva becomes tic chrutvd d]q qooIooJ*. These 
changes were no doubt carried out with absolute consistency. 
But as sandhi was of no use grammatically, it has been dis* 
carded in the modem Indian languages, as also generally in 
the other Arian languages, which in their earliest stages still 
show traces of it. But even in the present English we have 
such variations as -^ nuen, -^i 99^, hi9 -^ei aa, hkr -ij iz 

41. Natural speech is incessantly changing, both as regards 
its phonetic and its logical structure. The child learns the 
sounds of its vernacular language by a process of slow and 
laborious imitation. This imitation is always defective. If 
the child has been carelessly taught, or if it suffers from 
intellectual or organic defects, the divergence of its sounds 
from that of its parents may be so marked as to make its 
speech unintelligible to outsiders. But even under the most 
favourable conditions there is some divergence, for it is 
impossible for the child to reproduce by mere imitation the 
exact organic movements of its teachers. Even when the 
individual has settled down to a definite sound-system of 
his own, he is still liable to modify his sounds from laziness 
and carelessness. Even if the changes thus produced in the 
transmission of a language from one generation to another 
were imperceptible to the ear, their repetition would be 
enough to account for the most violent changes, if we only 
allow time enough. 

42. Hence we see that, as a general rule, all sound-change is 
gradual: there are no* sudden leaps in the phonetic history 
of a language. Such a change as the frequent one of it into 
ai presupposes a number of intermediate stages : X«, XtI, [x, |)r, 
\£, ]i, etc. Hence also there are no simultaneous changes of a 
sound, only successive ones. Thus we cannot suppose a simul- 
taneous opening and unvoicing of m, but only some such series 
as r, 3J, 3, 0. 

48. The sound-changes carried out within each language 



1 6 HISTORY OF ENGLISH SOUNDS. 

are uniform. This is the result partly of the tendency of 
the same mis-hearings and mis-reproductions of sounds to 
repeat themselves spontaneously in the pronunciations of 
most of the individuals of a community, but mainly of the 
social development of language, which tends to get rid of 
those new pronunciations which are in the minority. Even 
if two different organic tendencies were equally developed 
— even if one half of the children of a community mis- 
pronounced v/ as o, the other half as s^there would always 
be other considerations, such as distinctness, which, however 
trifling, would be enough to turn the balance. 

44. But the consistent carrying-out of a sound-change 
does not necessarily imply that it is carried out every- 
where, regardless of its position in the breath-group, its sur- 
roundings, and the influence of synthetic elements : quantity, 
stress, and intonation. On the contrary, most sound-changes 
seem to begin under special circumstances, and if they do 
extend themselves over the whole range of the sound in ques- 
tion, it is only gradually. A change such as that otwd into 
o t may begin at the end of a breath-group, and be then 
extended to the ends of words within a breath-group, as 
in German, and finally to all the ids in the language, as when 
every Arian d become a ^ in Germanic. Another change may 
begin in unstressed words, enclitics, etc. ; thus the E. change 
of ] a into the i of man is partially carried out in the Swedish 
a in unstrest syllables, which is ]k 

45. One result of the variation of change according to the 
stress is the formation of doublets, such as E. (ksen) and (kon) 
=^can,' the weai or unemphatic (kon) being the regular re- 
presentative of the strong (kaen) when unstrest. Here weak 
coincides with unstrest. But it often happens that an originally 
strest strong form comes to occur unstrest also ; thus the strong 
(hsev) is used both strest and unstrest, but with a difference of 
meaning in such a sentence as (-juwl hsev -to -hsev -jo heo 
:kat), the weak (-hov, -ev) being used only as an auxiliary. 
Originally weak forms often come to be strest. Thus (wilS) 
was originally the unstrest form of (wi)?), but it has now sup- 
planted the latter entirely. These changes of usage are partly 



SOUND-CHANGE. 1 7 

the result of divergence of meaning between the two members 
of the pair. Thus (ov) was originally the unstrest form of 
(of), but the two are now felt as independent words, and ov 
has developed a new weak form (ev) or (e). 

Doublets may arise in other ways as welL Thus in E. no 
when uttered in a deferential or conciliatory tone tends to r^^i), 
when uttered with decision or dogmatism it remains 7}9-). 

46. Sound-changes fall naturally under two main classes, 
internal and extemaL Internal changes, or sound-changes 
proper, are due either to the tendencies of the organs of speech 
themselves, as when (ii) becomes (ai), or to the acoustic quali-^ 
ties of the sounds themselves, as when / is substituted for y 
by defective imitation. We have, therefore, the subdivisions 
organic and acomiic, the latter often running directly counter 
to the former. External changes are quite independent of 
the nature of the sounds themselves, and are, as their 
name indicates, due to external causes, generally, but not 
always, connected with the expression of ideas. Thus, to 
take a familiar example, the change of asparagus into sparrow^ 
grass is due entirely to the attempt to substitute familiar for 
unfamiliar sound-groups, and a significant for an unmeaning 
whole. External changes are often quite opposed to organio 
tendencies, but they are essentially connected with acoustic 
change, for they always imply a certain similarity in sound 
between the old and the new form. It is, therefore, possible 
to include acoustic and external under the common head of 
inorganic, thus substituting organic and inorganic for internal 
and external as the primary divisions. 

47. Another important distinction is that of isolative and 
combinative. Isolative changes are those which afiect a sound 
without any reference to its surroundings, while combinative 
changes imply two sounds in juxtaposition, which modify 
each other in various ways. 

48. The consideration of sound-changes naturally includes 
the negative phenomenon of loss. The addition of a sound is 
generally only apparent — due to the practical exigencies of 
phonetic notation. The change of at into haty for instance, is 
merely a case of shifting of force (25). 

c 



1 8 HISTOEY OF ENGLISH SOUNDS. 

49. Organic changes fall further under three main heads : 
(i) tAroat-ohejige&i especially those which produce the im- 
portant fundamental distinctions of breath and voice; {2) 
changes oiform (stop to open, etc.) ; (3) changes of place (back 
to firont, etc*). 

INTERNAL ISOLATIVE. 

Breath and Voice. 

50. The relations of breath and voice in consonants are 
mainly determined by their surroundings, as when t between 
vowels becomes d, and consequently fall under the head of 
combinative changes. It is, therefore, difficult to determine 
whether the tendency of consonants, apart from assimilative 
influences, is towards voice or breath. The only absolutely 
unmistakeable cases of isolative change between breath and 
voice are those which run counter to the principle of assimi- 
lation, namely those in which a voiced consonant flanked by 
vowels becomes voiceless, as has happened in the case of the 
stops, both in the first and second Germanic consonant-shift (OE 
eian^Iui edere^ Qm leitefi=OE ladan), and in that of all con- 
sonants in many Middle and South German dialects. Change 
from voice to breath is easier initially and finally, and is very 
common in the latter case. In German, Dutch, and Russian 
all final buzzes and stops are unvoiced, although Dutch still 
voices final s in stressless words such as is and was when a 
vowel follows. This is evidently a tradition of the more 
primitive Sanskrit usage, which devocalizes finally only before 
a pause or a breath consonant. Liquids are rarely unvoiced, 
as in Welsh initial //- C3\, rh- oio. The evidence is plainly in 
favour of the natural isolative tendency being to change 
voice into breath. If we consider' that a voice consonant such 
as d is really (t) + (0), we see that the change of d into t is 
really equivalent to dropping a final obscure vowel. The 
tendency to unvoicing is shown most strongly in the stops. 
The explanation is that the stops are voiced with greater 
difficulty than the open consonants, the voiced breath having 
to be driven into an air-tight chamber, so that a voiced stop 



SOUND-CHANGE. 1 9 

cannot be held for any length of time. Liquids and nasals 
are not often unvoiced, because their audibility depends 
mainly on their sonorous vocality. But even vowels are 
occasionally unvoiced, especially the consonant-like highs, 
when final, especially after a breath stop, as in French vScu 
>[Q'f:, Russian ruki CDila'\I:. 

61. The intermediate change of voice into whisper is very 
common. Even English finds it easier than voice in such 
words as raged (oDrcDeiioi. In Portuguese final unstrest vowels 
are often whispered, as in campo Q'X^Db. 

52. The converse change of breath to voice always seems to 
be combinative, though it is sometimes extended by analogy 
to the initial and final positions, as in Danish skib * ship ' (now 
sqX«x^), due to the analogy of the medial b of skibe^ etc. 

Vowels. 

63. Narrow and Wide. As regards narrow and wide, short 
and long vowels follow directly opposite tendencies, short 
vowels being generally widened, long vowels narrowed, 
whence the pairs I«, I (ii, i), b, 1 (uu, a) in Qm, Icelandic, etc., 
as also in (Northern) English, L and 1 being apparently the 
older sounds of the short vowels. The change of a long wide 
into a narrow is shown in the Dan. vide >I«w\'[=MnIcel. >Jr4o] 
from older vita ^iJo!\, oId3, and in the history of the E. vowel 
in name. An example of a narrow short vowel is the E. ] 
in but, 

54. The high, consonant-like vowels L and 1 are liable to 
lose their syllabic value in juxtaposition with other vowels. 
This means of avoiding hiatus is a regular law of the Romance 
languages, where such words as gldnu soon became dissyllabic 
— aoo}*a>i3, which was practically equivalent to glarja. 

66. Place. As regards height, short vowels tend to lower- 
ing, as in Italian neve [ from nivem^ Dutch schip sciItD ' ship,' 
long to raising, as in E. goody Dutch goed cilo from older god 
G}4m, E. stone &om older (stoon). 

56. To this latter rule there seem to be no exceptions. 
There are, on the other hand, some cases of raising of short 

c 2 



20 HISTOEY OP ENGLISH SOUNDS. 

vowels, as in the change of e and o into % and u reap, in Gothic. 
Unstrest short vowels are often raised, as in Portuguese, where 
unstrest e and o become I and 1 resp. ; as in que^ campo, 

67. The tendency is from back to front. The frequent 
change of 1 into f , as in French lune^ Dutch zuur ' sour,' 
Was no doubt through the I of Norse htia and the !•• of 
Swedish hu8. The equally frequent change of 3 i^^ l> &s 
in E. fnan^ seems on the other hand to have been first to 3n 
and then direct into x» which only requires a slight forward 
shift of the configurative narrowing. 

68. But the front x occasionally changes into the back ], as 
in E. f(Mt from OEfast through (fsesest), probably in order to 
avoid confusion with the ^-sounds. 

60. The change from front to mixed is mainly in unstressed 
vowels, as in the Qm, Dutch, etc. -^=1, Portuguese que aT. 

60. Bounding. As regards rounding, back and front vowels 
follow opposite tendencies, back vowels favouring rounding, 
front unrounding. 

The first stage in the rounding of back vowels is forming 
them with imperfect mouth-opening, the low-back j and 
J being especially liable to this muffling. Indeed, unless 
pronounced with very open mouth, these sounds are always 
apt to be mistaken for rounded vowels. But muffled, or even 
fully front-rounded j) is still distinct from J= J^. The round- 
ing of d into some variety of open o is very general in the 
Germamc languages. 

eL The unrounding of back round vowels is rare. We see 
partial unrounding in the short Swedish 7^=1», complete in 
the English u vnbut ^'l. In unstrest syllables the change is 
commoner, as in OE boga from older bogo. Unrounding of 
front vowels is shown in the later OE fet for earlier fat, in 
MnE 9in from OE si/nn sf T«, and in the South German pro- 
nunciation of ii and o 9^ i and e. Partial unrounding in 
Swed. y=f), Im, distinct from French «=f. Of the rounding 
of fronts I have no examples to hand. 

62. Examples have already been given of under-rounding 
(6x). Of the other kind of abnormal rounding, nl over-round- 
ing, examples are afforded by the Swedish and Danish j)«, as in 



SOUND-CHANGE. 2 1 

9^^ ^ go/ and })4, BAm9oV sun/ which are special Scandinavian 
modifications of j« (from a) and }4. 

63. Diphthonging. Isolative diphthonging or 'vowel- 
cleaving ' mainly afiects long vowels, evidently because of the 
difficulty of prolonging the same position without change. 
Cleaving of high vowels, as in the very frequent develop- 
ment of (ai) and (au) out of (ii) and (uu) resp., begins with a 
slight lowering of the first half of the I* or !♦, giving XtX or 
I-rf, as in Southern £. mcy which is practically almost equiva- 
lent to (ij). !♦ and !♦, however, are more easily cleft by simply 
increasing the lip-narrowing towards the end of the vowel, so 
as to form a consonant, as in the Southern £• who slo. 

64. Mid and low vowels are cleft by a slight raising of the 
tongue ; or, in the case of round vowels, by a progressive 
narrowing of the lip-opening, which may, of course, be accom- 
panied by a raising of the tongue. Examples are the English 
Dtt and }9-) in my and no. In the latter the cleaving is effected 
entirely by the lip-narrowing. 

66. All of these are falling diphthongs. Rising clefts are 
the Italian x(J, ^ irom Latin ^, 9 through [«, }«, as in lieve^ buono 
from levisy bonui. In Mnlcel. e is regularly cleft into o>||«, as in 
mer * me.' 

66. But diphthongs may also arise from lowering the second 
half of a long vowel. In North Welsh all long high vowels 
are followed by an obscure voice-glide : X#i, i#i, I«, as in drtoi^ 
Such was probably the beginning of the Old German no from. 
^, as in muot. 

It will be convenient to discuss all the changes of diph- 
thongs under the present head, although some of them fall 
under that of combinative and acoustic changes. 

67. In diphthongs of the (ij)-type there is a tendency to 
make the cleaving more distinct to the ear by divergence, the 
first element being lowered and retracted through [, [, X, 3, j^ 
etc., or even rounded, as in the Cockney pronunciation of mjf 
as fjIt. (uw) is diverged by lowering and unrounding — Ja-, J9-, 
J9-, and then by fronting as in the Cockney now Tp. Diph- 
thongs beginning with a front-round vowel are diverged by 
backing this front vowel, as in the Danish oie^ now=:|xX. 



22 HISTORY OF ENGLISH SOUNDS. 

68. While the street element shows this tendency to diverge, 
the glide shows the opposite one of approximation, ]£, for 
instance, becoming ]£t, 3g, 3^, till at last the diphthongic cha- 
racter is almost lost, as in the Cockney pj«ico=^?7^, almost 
indistinguishable from marie. 

09. If the glide-vowel is fully formed, it often acts on the 
preceding vowel as in mutation (143), drawing it towards 
itself, so that ji, 33: become i£, (Jx, and ^1 resp., a;^ in Olcel. ei 
from ai, gu from au in s^ein, guga (auga). 

70. * Smoothing,' or the levelling of the two elements of a 
diphthong under a monophthong, is the result of absorption, 
as when ai becomes d in OE 9tdn by lowering of the glide, ei 
becomes ^, as in the Swed. and Dan. den. This direct absorp- 
tion is, of course, only possible after considerable convergence 
of the two elements. 

71. Forward smoothing, as in Germanic change of ei into I 
in wln^ is only the completion of the mutative influences 
described above. 

72. As cleaving is peculiar to long vowels, it follows that 
when a diphthong is shortened, as when it stands in an 
unstrest syllable or before two consonants, it tends to smooth- 
ing, as in Icel. eld from *eild. 

78. Of course, it is possible to make the glide-element of a 
diphthong so short that the whole combination can be regarded 
as the equivalent of a short vowel, as in OE ea^ eo. 

14t. Another result of the strengthening of the glide is that 
it sometimes develops into a consonant, as in Mod. Greek 
amos }>'0}s. This development is the rule in most languages 
in rising diphthongs, ia generally becoming ^a. 

76. Loss. Isolative loss of vowels seems to occur only in 
unstrest syllables. Even here it is possible that the loss is 
only apparent, being compensated by lengthening of the pre- 
ceding sound : we may perhaps assume that Middle E. name 
became 'ijir* as the first stage of its present monosyllabic 
form. 

76. The dropping of unstrest vowels is generally preceded 
by various weakenings, generally in the direction of V Drop- 
ping without previous weakening is, however, common in 



SOUND-CHANGE. 23 

spoken Welsh, as in agonoch Q^iAc, But even in Welsh it is 
the mixed \ which is oftenest dropt, as in xfory >}(0l. 

77. The dropping of unstrest vowels is often dependant on 
the nature of the resulting consonant-group. Such combina- 
tions as k{9)l^ ^(^)», in which the second element is a vowellike 
syUabic, are especially liable to contraction, especially when, 
as in the second instance, the two consonants are formed in the 
same place. But in Old Icel. we find vowels dropt without 
any regard to the nature of the resulting consonant-groups, as 
in l4ix gen. sg. from flaxes through *laiss. 

78. The contraction of two short vowels into one long, 
which is a frequent means of avoiding hiatus, as in Sanskrit 
afi va^ati iva, implies, of course, only the loss of the inde- 
pendent stress with,which the second vowel begins. Where 
one, or both, of the vowels is long already, the contraction 
was no doubt at first extra long. 

Consonants. 

70. Form. The opening of stops generally seems to begin 
between vowels, and is then evidently due, in part at least, to 
the attempt to assimilate the form of the consonant to that of 
a vowel. This is confirmed by the fact that it is generally 
voiced stops that are opened in this way. Thus in Modem 
Qreek ^ has everywhere become e, while k continues to be a 
stop, and so with the other stops, the change having probably 
begun between vowels, and been then extended to the initial 
and final occurrences of the voiced stop. In Dutch too y 
has everywhere become ei or ci, while in German initial 
ff retains its stopped quality. 

80. But voiceless stops are sometimes weakened into open 
breaths between vowels, as in the regular change of c and t 
into cA and lA in Old Irish, as in atAir. In Danish unstrest -€6 
becomes -^vn, as in Auset. Here the change was probably 
direct, but in other cases it may be the result of strengthening 
the breath-glide (140). The frequent change of At^ pt into CQ, 
X7 seems to be partly due to striving after distinctness, as 
also that of it into 9t^ as in Latin eguealer. 



24 HISTORY OF ENGLISH SOUNDS. 

81. The change of a nasal into an open consonant is, of 
course, through a nasalized open consonant ; thus the Welsh 
mutation of tnam into fam >]r postulates an intermediate 953^, 
which is nothing but an m with the lip-passage open. 

82. The change of a stop into a side-consonant is not com- 
mon, but there are examples in Latin, such as oleo by the side 
of odor, 

88. Side-consonants are capable of a further weakening 
into open consonants, as in French fille/ milieuy where 00 has 
become O), liA^BJifamma from fiamma, 

84. The change of an open consonant into a vowel is, in the 
case of y and w^ often almost entirely dependant on stress- 
shifting and synthesis. In OE mdw^ for instance, it is im- 
possible to tell whether the w really m^ans 8 or is simply 
equivalent to l ; most probably the latter, but the distinction 
is very slight. 

85. Some consonants, such as E. and Gm r, are pronounced 
with such a complete absence of buzz and with so open a 
configurative paasage that they may be regarded hb gUde- 
vowels rather than consonants. 

86. These weakened consonants must be carefully distin- 
guished from syllabic consonants. It is true that the un- 
buzzed vowellikes lend themselves with especial ease to the 
syllabic function, and that it is possible that the Sanskrit 
vocalic r in mrtd really had something of the E. r in it, but 
there is nothing to prevent it from having been a strong trill 
— at least at first. 

87. If the configurative passage of an open breath con- 
sonant is progressively enlarged, the acoustic effect of its 
position becomes more and more indistinct, till at last we 
hear nothing but mere breath. In modem Irish the old th^ 
* aspirated ' «, etc., are weakened in this way to mere As. In 
Sanskrit final a becomes a mere breath, as in dgwah. Even in 
E. / think often sounds like I hink. 

88. All these changes are weakenings. The converse 
change from open to stop is fr^uent. The open voiced 
consonants between vowels are especially liable to this 
change. Indeed such a consonant as e, if pronounced with- 



SOUND-CHANGE. 25 

out perceptible buzz, as in Mnlcel. saga s]«e], has very much 
the effect of Q. The Old Swedish mgha has, accordingly, 
become sj«g] in MnSwedish. (d has been stopped in Greek 
zug6n:=^msi€!^'}y Latin jugum^ and in Italian gia (De3 from jam^ 
all pointing to a preliminary stopping of O) into m. In 
Swedish such a word as Jag, when emphasized, is often uttered 
with weak stoppage^ so as to be intermediate between <dj4Q 
and flg-^Q. The change of 9 into Q through G is seen in the 
French garde from Qerman warda, Welsh gtolad Qm^m = OLrish 
flailA (where /= older w), showing the intermediate stage a». 
A parallel change of the voiceless o is seen in the dialectal 
Icel. pronunciation of Avalir as q>3«coIcd. The stopping of u, w 
is common to most of the Germanic languages, as in Swedish 
ting, German dingzsz'E. thing, Swed. dumb^^-^. thou. 

89. We have, lastly, the trilling of open consonants, espe- 
cially (D and e. The tendency of the dialects of large capitals 
is in favour of untrilling, as we see by comparing the London 
with the Edinburgh, the educated North Gm with the pro- 
vincial and the Dutch r. Dutch, on the other hand, not only 
retains a strong (01, but also trills its ^=€i, and its c in schip:^ 
SCiItD. Trilling is no doubt the result of striving after 
distinctness. 

80. Plaoe. Jiack to Throat. We see this change in the 
Danish rsse^, which is no doubt a later form of the €i of the 
Jutland pronunciation. In Glasgow Scotch t in butter, etc., is 
en — ^a t with simultaneous glottal stop. 

81. £ack to Front. This change appears to be always com- 
binative. 

82. Front to Back. Italian valga from valeam through 
^Jq^] and '^Jooo]. So also It. tengo from teneo. 

88. Forward to Back. The frequent substitution of ei for (Df, 
as in the Parisian r, seems to be mainly imitative. For 
Russian /, see § 104. The change of a into c, as in the Old 
Bulgarian choditt, seems to be the result of inner rounding 
and subsequent exaggeration of the back element, as in the 
change of w into gw (88). In Armenian we find sw developing 
into a back aspirate stop through c, as in khuir ' sister.' The 
first stage is shown in the Gm sch^ty In the South Swedish 



26 HISTOBY OF ENGLISH SOUNDS. 

pronunciation of z in iijuta, etc., the inner rounding is very 
marked, the point of the tongue being lowered, which would 
soon develop back modification. 

04. Forward to Front. Spanish l from Latin nn, as in aho. 
So also in some West-Norwegian dialects II and nn become 
€X0 and ll, or approximations to them. 

96. Lip to Lip^teeth. In the change of jo into^J w into v, we 
may always assume an intermediate o, 3, the latter being the 
Middle German w. This is partly an acoustic change, > being 
more distinct than o. 

The converse change is shown in the Danish hav ^ 8ea'=s3». 

96. Forward to Lip {-Teeth). The frequent change of w to >, 
as in a defective pronunciation of through, and in laXm/utfius 
= Sanskrit dhumds, seems to be imitative, but may sometimes 
be through V). 

97. The converse change is shown in that of final m into », 
as in Spanish Addn. 

98. There are various changes of the forward consonants 
among themselves. That of (z) into (r) is frequent in Latin, 
as in Aurora, and in Germanic, where it was through (D^ (i45}- 
The converse change is shown in the older Parisian Fr chaise 
from chaire. 

99. Isolative change of S into z is regular in Gm initial s 
followed by a cons., as in schwan, stein. In Portuguese s final 
or before a consonant becomes the intermediate e^, as in casas 

100. Liversion is generally the result of the influence of r, 
a sound which always tends that way, especially when trilled — 
as in Swed. bam Dj*Tc. The inverteds are strongly represented 
in Sanskrit under the name of * cerebrals,' where they are pro- 
duced by the influence not only of r and J, but of other 
sounds, even the vowels i and u. It is possible that inversion 
may in some languages be the result of exaggerating the 
distinction between gum-point, such as E. t, and teeth-point 
conss. 

101. A peculiar result of inversion is the change of I into 
the Scandinavian * thick I' (33). 

102. As regards rounding, back open consonants tend, like 



SODND-CUANGE. 2 ^ 

back vowels, to rounding, as shown in the history of such 
words as draw from OE dragan^ sorrow from 9org. E. r is 
rounded in individual pronunciation. The rounding of 9 and 
/ has been treated of above (93). 

The tendency of back-round conss. to exaggerate the back 
element has also been illustrated above (93). 

108. The loss of back modification is shown in the frequent 
change of (w) into (v) through 3, as in Gm. 

104. The most unstable of the conss., as regards modifi- 
cation without change of place, is /, whose position can be 
combined with that of almost any vowel. In the ' clear ' I the 
front of the tongue is somewhat raised in the direction of oo\ 
which gives the French /. In the * dull ' English / the front 
of the tongue is hollowed out. The Dutch / is decidedly 
back-modified or ^ guttural,' still more so the Portuguese / in 
altOy which is quite cd(. The clear / tends to become oa, the 
dull to become 8, as in Bussian ji^a/^a. 

106. Cleaving. Consonant-cleaving, as when U becomes dl^ 
nn becomes dn^ as in Mnlcelandic /a/^ ^J^'^S) ^^^ [^"^9 is* 
like vowel-cleaving, the result of the diflSlculty of prolonging 
a consonant unchanged. In the West Norwegian dialects the 
dl in falla is articulated so lightly that the combination is 
really half-way between U and dl. Another kind of cleaving 
is shown in tiie Welsh nh in nhad=i2^m, which must once 
have been simple i]«(D. 

106. Consonant-smoothing is analogous to that of vowels. 
It is forwards in Danish hinde dIiI through *DXn\, backwards 
in German jo/fe^<?r from *pfepfar. 

107. ItOBS. Consonants are more fr'eely dropped than vowels, 
as being less sonorous. Thus Germanic initial j is dropped 
everywhere in the Scandinavian languages, as in Icelandic dr 
^year. The loss of h in Cockney and provincial K is only 
apparent, being due simply to a shifting of force (25}. The 
dropping of initial k in know was preceded by a stage in 
which it unvoiced the », so that the k was only dropped 
beoatise it had become superfluous for distinctive purposes. 
Many other consonant-droppings are no doubt due to the 
same principle of economy of distinction. Final consonants 



28 HISTORY OP ENGLISH SOUNDS. 

are very easily dropped, being uttered with less force than 
initial ones (17). The audibility of final stops depends mainly 
on the off-glide, and if this is suppressed, they become almost 
inaudible, and this was probably the beginning of that whole- 
sale dropping of final conss. which we see in French. French 
keeps final conss. before another word beginning with a 
vowel (liaison), but Old Bulgarian drops all final conss. with- 
out exception, nasals alone partially surviving in the nasaliza- 
tion of a preceding vowel, so that every word in the language 
ends in a vowel. Other languages, such as Greek, allow only 
certain consonants at the end of a word — mostly vowellikes. 
Final consonant-groups are often very deficient in sonority, 
especially stop-groups, such as kt^ and are consequently light- 
ened by throwing off the last, as in the Cockney pronuncia- 
tion of act as \cxk ; other groups being lightened in various 
ways, as in Greek dnax for *dnakt9. Even polite £. makes 
asked into (aast). 

108. Addition. The addition of d and p in such combina- 
tions as an{d)ra, am{p)ta, as also their dropping, is only 
apparent. In passing from n to r it is necessary to shut the 
nose passage, and open the mouth passage simultaneously, 
and the slightest delay in doing the latter of course converts 
part of the r into a d. 

109. The addition of hiatus-fiUing consonants, as in Dutch 
zeeen^^s^l^^Xlf is simply due to a slight exaggeration of the 
glide between the two vowels. Such insertions as that of r 
in the £. idear of are, of course, the result of external analogy. 
Other means of avoiding hiatus are the glottal stop — ]x], and 
the introduction of a breath-glide, as in tlie occasional French 

An interesting example of consistent hiatus-filling is afforded 
by Old Bulgarian. In this language, as already stated, every 
word ends in a vowel. So, in order to get rid of all hiatus, 
every initial vowel developes an allied consonant before it, i- 
becomingy»-, #- becoming vii (from *ioU-), etc. 



SOUND-CHANGE. 29 

Quantity. 

110. There is a general tendency to shortening in un- 
strest syllables, -aan, -ann both becoming -a«. In strest 
syllables there is a tendency to alternate short and long in 
vowel + cons. Final -an is often lengthened to -anny though 
the short cons, is kept in many languages, and final -aann 
generaUy becomes -aan. Medially ana tends to become aana^ 
aanna also to become aana. The frequent change of anna into 
ana^ as in Gm gewinnen^ seems to be the result of the quali- 
tative divergence of short and long vowels : when I4, 1 had 
been separated into I4, X, the doubling of the (n) in gewinnen 
became superfluous and it was therefore shortened. 

111. In the Romance languages stress keeps a final vowel 
shorty as in French n si^ while in the Germanic languages it 
lengthens. 

112. In many languages the high vowels, especially i and u^ 
tend to shortness, either resisting lengthening influences, as 
in E. 9(m^ toriUen from 0£ 9unUy writen^ or else being short- 
ened against the analogy of the other vowels, as in Dutch 
lieden cdIoi, voeten ^ioi. The extreme closeness of these 
vowels seems to make their lengthening difficult. 

118. Shifting of quantity often accompanies stress shifting 
in diphthongs, as in Olcelandic kjom from *ieosa. 

114. Vowel quantity is often dependant on the influence of 
the following consonants. Stops, especially voiceless stops, 
shorten. The shortening influence of m in Swedish is very 
marked. In Welsh j^, m, / shorten, often also n. r lengthens 
in many languages. 

116. Vowellikes and nasals followed by another consonant, 
especially a voice stop, often lengthen, as in E. beard, mid, 
fnd. The lengthening seems to be due to the difficulty of 
distinguishing the vocality of the vowel from that of the 
vowellike, (finnd) and (fiind), for instance, having much the 
same effect on the ear. In some cases the lengthening of the 
vowel is due to the absorption of a parasite- vowel (159), as in 
E. (haad) ^ hard' from (hard) through (haard). ' 

116. The distinctions of quantity are utilised differently in 



30 HISTORY OF ENGLISH SOUNDS. 

different languages. In many languages, such as Bussian, 
French, and the Bomance languages generally, the distinction 
of long and short vowel is not clearly marked^ the quantity, 
especially of stressed vowels, being generally medial. Other 
languages, such as E., distinguish accurately three degrees of 
quantity. Again, in such languages as Sanskrit and Hun- 
garian any vowel in any part of the word, whether street or 
not, may be long or short ; but in other languages quantity 
is partly dependant on stress and position. In Welsh the 
last syllable but one is stresst and short, so that ton and Ion 
both have plural tonau d^l^. In Swedish such a word as 
E. biUer with a short strest vowel followed by a short cons, 
is impossible. In E. such a word as Qerman mann p]*?, with a 
short final cons, preceded by a short strest vowel, is equally 
impossible. 

117. The influence of quantity on other changes is very 
marked, especially as regards vowels. Long vowels tend to 
narrowness, raising, and cleaving ; short vowels to wideness, 
lowering, and smoothing. It also influences stress, as in 
Latin (i2i). 

Force. 

118. Stress-shifting in diphthongs does not affect those of 
the (ai)-type, but only when a closer (higher) is followed by 
an opener (lower) vowel, or a back by a front. Whenever the 
first element of a falling diphthong by gradual divergence 
reaches (i) or (u), as when x* passes through (Ji, ()t into f j, 
there is a tendency to shift the stress on to the opener and 
more specifically vocalic second element, a tendency which is 
no doubt helped by the difliculty of lengthening a high vowel 
(112) and the ease with which such a vowel passes into a 
glide and a consonant. The two extremes are therefore the 
falling (*ai), and the rising (i'a) almost =(ja). Italian ie^ uo 
were no doubt originally falling diphthongs, (u) is felt as 
opener than (i), hence (iu) tends to (i'u, ju). In the South 
Gm dialects no from a through (oo, oa, ua) still remains a 
rising diphthong, as also ie from io. 



SOUND-CHANGE. 3 1 

110. The general tendency of language is to alternate strong 
and weak strest syllables as much as possible. Hence the 
tendency to throw forward the stress of a two-syllable en- 
clitic in some languages, as in Old Icelandic, where Jmr era 
becomes ^Wr to, 

120. Some languages, such as French, have practically no 
independent stress, intonation taking its place. 

121. In some languages, such as Sanskrit, Bussian, and 
English, the place of the stress may be on any syllable in the 
word. In others it is fixed on some one syllable, as in 
Welsh, where it is regularly on th« last but one, and in Ice- 
landic, where it is on the first. In others, again, its place is 
partly determined by quantity. Thus, in Latin it must be on 
the last but one if that syllable is long, on the one before that 
(third from end) if the last but one is long : mowere^ 'regere. 

122. Stress-shifting in different syllables is due partly to 
such mechanical limitations, partly to external influences, as 
when a language throws the stress on to the root-syllable. 
This may often be effected most imperceptibly, by gradually 
increasing the strength of an orginally only half-strong 
syllable. 

128. It must be remembered that originally stress was due 
to purely external causes. Here we may observe two opposed 
tenc^encies: (i) to emphasize the most important element of a 
word or group, as in ' a piece^ of 'bread' ; {%) to emphasize the 
element which modifies the original meaning of the word to 
which it is added, as in ^ to give and 'forgive.' The first 
tendency leads to putting the stress on the root, the second to 
putting it on inflections, etc. 

124. The influence of stress on sound-changes in general 
is very important. All the weakening processes, shortening, 
dropping, assimilation, smoothing of diphthongs, etc., begin in 
unstrest syllables. Thus Icel. skgpujm from ^skapafiu shows 
only approximation to the u (mutation) in the strest syllable, 
but complete assimilation in the second, unstrest syllable. 
As the beginning of a syllable generally has the strongest 
stress, initial and medial consonants often show the opposite 
tendencies of strengthening and weakening, as in Danish kage 



32 HISTORY OF ENGLISH SOUNDS. 

0^3^*-! =OTcel. kaka^ where the first k is strengthened, the 
second weakened. 

125. The shifting of a cons, from the end of one word to 
the beginning of another, or vice- versa, as in nickname from an 
ekename, adder from a nadder, is really due to shifting of force. 
So also is the loss or addition of A. 



IntonatioiL 

126. Intonation was orifpnally an instinctive means of em- 
phasis, an energetic utterance of a vowel being accompanied 
by a high tone — ^level or rising, the unemphatic syllables 
being uttered in a low tone. 

127. Hence the intonation in primitive languages —at least 
in Sanskrit, Qreek, and the other OArian languages — is 
fixed in each word : it is a toord-intonaHon, This fixed in- 
tonation still survives in Lithuanian and Swedish. In 
Swedish, for instance, j>«d})4 uttered with a rising tone is the 
town Abo, but if uttered with a falling tone on the first syl- 
lable and an upwai*d leap on the second, it means ' dweller.' 
Even in E. Vaa'Sp and vaai9 have the contrary meanings of 
' a little ' and ' very much,' as in answer to the question ' does 
it rain?' 

128. In the more highly developed living Arian languages, 
on the other hand, the intonation is not bound to any one 
syllable of a word, but is used to modify the meaning of the 
sentence as a whole, a rising tone implying doubt, question, 
incomplete statement, etc., a falling tone certainty, answer, 
completion, etc. Even in Sanskrit and Greek the word-tone 
was no doubt modified by these tendencies, as it certainly is, 
not only in Swedish but also in Chinese — a language in which 
word-intonation plays an exceptionally important part. 

120. Intonation is not necessarily associated with stress, 
but there is a strong natural connection between them, and 
the history of the Arian languages shows clearly that in them 
high tone was accompanied with strong stress, for the weak- 
ening and dropping of vowels in unemphatic syllables which 



SOUND-CHANGE. 33 

is carried to such an extent in parent Arian cannot be ex« 
plained as due to mere lowering of tone. 

ISO. There is, therefore, no such thing as a change of 
modulative or 'musical' accent into stress-accent: all that 
has happened — say in Modem Greek — is that the stress has 
been kept, while the intonation has been set free. 

181. The compound tones are often accompanied by double 
stress on the intoned vowel (zweigipfliger accent) which seems 
to cut it into two. This may be, as suggested by Sievers, a 
cause of diphthonging. 

182. The Danish substitution of the glottal stop for the 
' simple ' intonation of Swedish, as, in p3xi» mand = Sw. mann 
^rj'74, is very remarkable. It is evidently due to an energizing 
of the intonation. Even in Sw. the simple tone is often ener- 
gized in such words as baron, the vowel being pronounced 
with a jerk in the middle so that it seems to be divided into 
two, a falling being at the same time substituted for the rising 
tone — dX^^})})*?. In some Lithuanian dialects (according to 
Eurschat) the same thing happens, which in Lettish seems to 
develop into a full glottal stop, as in Danish. 

188. The influence of intonation on sound-change is very 
slight. It seems, however, that in parent Arian a with a 
high ton^ became «, while a low tone changed it to 0, evidently 
because e has naturally a high, a low pitch. 

184. As regards the relation of intonation to quantity and 
stress, we may say briefly that emphasis, length, strong 
stress and high pitch are naturally, though not necessarily, 
associated. 

Transposition. 

186. Transposition, as in OE asnan for ascian, MnE bird=:OE 
bridd, is generally a more or less isolated phenomenon, but is 
sometimes carried out through a whole group of sounds, as 
when Greek z=iJDS became 8Q> in the Attic dialect. 



34 HISTORY OF ENGLISH SOUNDS. 



INTERNAL COMBINATIVE. 

136. The influence of one sound on anotiier may be either 

forwards, aa when adta becomes adda, or backwards^ as when 

adia becomes atta. It may be &\\heir partial (subsimilation), as 

in the Germanic vowel-mutation (j%**, /»**«), or complete 

(assimilation), as when adta becomes atta. 



Breath and Voice. 

187. The change of breath stops into voice between vowels 
is regular in Danish, where, however, the resulting voice stops 
have been opened, as in lade ^ let ' ooJi^vn]^, skibe SQbi5\ ' ships.' 
In Sanskrit final stops are always voiced if the next word 
begins with a vowel, as in udeti ' goes out.' The hisses, such 
as «, are not voiced either in Danish or Sanskrit between 
vowels. 

188. Breath and voice assimilation between two consonants 
is almost universal in language, such combinations as English 
lA in backboard being exceptional reactions due to striving after 
distinctness. In Dutch bakboord is pronounced Dj€iD}4<i){0, and 
when a Dutchman speaks English, he is apt to make Butch 
Jews into Budge Jews, Sanskrit follows the same laws, even s 
being voiced before voiced conss. and vowels, although in the 
extant language the resulting z has been changed to r, or 
dropped, as in dgwd dramatic where -J stands for ^-oz. Even 
E. has (ksets) against (dogz). 

130. It will be seen that the stops are the most sensitive to 
breath and voice assimilation, while the vowellikes r and I 
and the nasals are generally quite independent of them. In 
Icelandic, however, It, etc., as in bilt, becomes coo. 

140. The slight escape of breath which follows breath-stops 
in such languages as E. is easily developed into an aspirate- 
glide, as in Danish kan Q^^l^, Sanskrit kh, by a slight stress on 
it, which at the same time relieves the pressure involved in 
forming an unaspirated Q'. It is, therefore, a mistake to 
suppose that an aspirate requires greater effort than an 



SOUND-CHANGE. §5 

ordinary stop: there is simply a shifting of effort from the 
stop itself to the glide. If an aspirate-glide is held 
unchanged, it becomes a definite, open consonant corre- 
sponding to the preceding stop, giving the combinations 
known as ^affricates' or ^stop-opens': Qc, oo, DO (d:>). t 
developes in this way either into ou or OS, according as it is 
a pure point or a blade-modified stop. The glide in the 
Danish o^ in tale sounds between s and Jf, When the glide 
has thus obtained an independent existence as an open con- 
sonant, the stop itself is often dropped as superfluous, as in the 
German jifeff'er from dxdJcd through ty>io^](i>^ dox(d)o3(d. The 
front stop Q is peculiarly liable to these developments, its off- 
glide being very liable to develop into full o because of the 
difficulty of removing the broad ridge of the tongue quickly 
enough from the palate. Indeed even q' always suggests ^J 
to an unaccustomed ear, the glide being so distinct. In San- 
skrit the aspirate of c o, which is written c^, must have been 
really the stop-open oo, for it makes a preceding vowel ^ long 
by position.' 

The influence of i in aspirating an adjoining breath-stop is 
seen in Sanskrit stiitds, gdcchati qSqqoJoI and Qreek skMzo. 
In Danish it has the opposite effect : cp tU QofcD with atille 
so'Icd]^. Here the initial 9 seems to take away the stress from 
the following cons. 

141. The two chief kinds of influence of vowel on vowel are 
vowei^Aarmony and mutation (umlaut). Mutation, however, is 
backward and indirect, implying modification of the inter- 
vening consonant, while in vowel-harmony the influence is 
generally forwards, and the one vowel acts on the other 
directly without any necessary modification of the intervening 
consonants, and therefore extends more easily through an 
indefinite, number of syllables. It appears to be partly 
acoustic. The best example of it is in Finnish. In Finnish 

the vowels are : 

(i) hard : a, o, v. 

(2) soft : a, b', y. 

(3) neutral : «, i. 

In Finnish the root-syllable always comes first and has the 

D % 



36 HISTORY OP ENGLISH SOUNDS. 

chief stress. If the root-syllable has a- hard vowel, all the 
following must have a hard or (more rarely) a neutral vowel ; 
if it has a soft vowel, all the other vowels of the word must 
be soft or neutral : muuttumattamuudestama, tyytymMtdmyyd^ 
esldnsd. For traces of this in the Arian languages see § 159. 

Front-Modification. 

142. The influence of i and the other front vowels and of 
j on a preceding back cons., especially the stops k and ^, may 
be seen in any language. Even in E. the k in keen is a little 
more forward than in com. In Russian the front vowels 
I, [([), and the now silent ?(=X?), communicate their own 
front articulation to most preceding consonants, but in various 
degrees according to the nature of the consonant. (01 r, s «, 
>/, r m, op simply arch the tongue into the x-position 
(=©►), or, in other words, anticipate an x, but without 
otherwise modifying their original articulations. In such a 
group as Ijrd (i^^) the i-position is maintained unchanged 
from beginning to end. In such a word as miru rv[(Oi a 
Russian brings the tongue into the x-position simultaneously 
with the closure of the lips which forms the r. Q k becomes 
Q\ as in the old-fashioned E. kyard=Qy with simultaneous Qi. 
c (x) becomes o (q). The fronting is carried out most fully 
with the point nasals and stops i n, u t, m dy whose place 
of stoppage is shifted back to the outer front position, both 
cons, and vowel in li-I, qi-I being apparently formed in the 
same place, the point of the tongue not being employed at 
all. The fronted CD I was probably once oos but it has now 
become almost the ordinary point consonant, probably be- 
cause its wide divergence from the back 8 in palia made 
further differentiation superfluous, e J and e 3 have also lost 
their original front modification, at least in the Moscow pro- 
nunciation. The loss of original fronting has been very ex- 
tensive in Servian and the other South Slavqnic languages. 

(d(j) has exactly the same influence, being itself dropt, as in 
difjd ad'd]. 

148. These fronted consonants again in their turn influence 



SOUND-CHANGE. 37 

a precedisg sound. Thus the Q in iesH t[s^Q fronts the pre- 
oediDg s, and this sn again narrows the preceding vowel 
(which would otherwise be the wide (J) into [. ^o) and 3(^) 
followed by a fronted cons, are advanced towards the mixed 
positions — J-i-, ]k The second element of the diphthongs ei and 
ai has the siEune effect — [x, ]y£. 

144. It is certain that these vowel-changes are due entirely 
to the direct influence of the immediately following cons., 
for if that cons, is not fronted by a following cons., as 
sometimes happens, the vowel remains unmodified, as in 
iripki Q(Di\|jDa\I. 

146. In these Bussian changes we have the key to the 
Germanic vowel-mutation or ^ umlaut.' In most cases the 
fronting of the conss. which caused the mutation has been 
afterwards given up, as in E. end^ which must once have been 
p\arN(D3-. But it still survives in such words as E. bridge^ 
OE brycg^ from D<0laH(D}4. That the Germanic mutation may be 
due entirely to cons, influence is shown by the regular IceL 
mutation caused by the fronted r which arose from ;2r, as in 
eyra from *auzd through Jl<o\j. 

146. It need hardly be said that all vowel-mutation takes 
place very gradually : that between the f of brycg and original 
i there must have been li>, f , f i*. 

147. But a front or front-modified cons, may influence a 
preceding vowel in a different way, nl by exaggerating its 
on-glide into a diphthongic voweL Such a group as 3p^> 
indeed, always suggests aimi to an unaccustomed ear, being 
really equivalent to 3(x+f)I. We see the results of this 
diphthongic mutation in such forms as Greek kieino from 
*klenjdf French gloire from gloria through *glarja, 

148. Forward front influence of vowel on cons, is shown in 
Gm. ich Xo contrasting with ach ]c. This is the opposite of 
Russian, where ich retains the c of ach. 

149. But in Bussian a fronted cons, draws forward a fol- 
lowing vowel, so that sjo is s\\y sometimes almost s^{. Such 
& word as French Sue is in Bussian written 9ju = 8^lK Un- 
strestya in Bussian is often weakened into (d|) (through '^o]*-), 
as in Jadro 0)|!aJ'(i>3-. 



38 HISTORY OF ENGLISH SOUNDS. 



Back-Influence. 

150. Back-influence is shown in the Russian development 
of CD into 8 before back-vowels, as in palka dJ^qJ- 



Bounding. 

Bounding influence is parallel to front influence, though 
less extensive and important. 

161. In Bussian rounded vowels (all of which are back) 
communicate their back-round quality to preceding consonants. 
This is most marked with c, which becomes c before the two 
round vowels : cl, c}. In pAs the i^-quality is also distinctly 
heard in the body of the cons. In a^J, w} of course only the 
ofl'-glide is heard, which sounds like a half-suppressed 8, so 
that an unaccustomed ear is apt to hear qp} alternately as 
ko and kwo. Only back and lip conss. are rounded in this 
way. In Old Icelandic we have an u- and t^-mutation, as in 
mgnnum fjllljF from *mannum, gora a\f (d] from ^garwjan through 
^g^rwa 0\(A^{g)% Diphthongic f«-mutation is seen in Greek 
pa4T0S from ^parwos. 

162. Forward rounding by a vowel is seen in German aucK, 
jk: ; by a consonant in 0£ wudu from widu through modu. 

Nasalizing. 

168. Nothing is more common than the nasalizing influence 
of a nasal on a preceding vowel. Indeed, it is doubtful 
whether any language is entirely &ee from this influence. 
It is common in E., and is often strongly developed in 
German. There are various degrees of nasality ; thus French 
is stronger than Portuguese nasality, the uvula being lowered 
more. When the nasality of a vowel is clearly developed, 
there is a tendency to drop the following nasal consonant as 
superfluous, whenever this can be done without causing a 
hiatus, that is, when the nasal cons, is final, or stands before 
another cons. This was carried out with perfect regularity 



SOUND-CHANGE. 39 

in Old Bulgarian, as also in French and Portugaese. One 
result of this is that in all of these languages d is wanting : 
Port. lon^o=co}iQh. 

154. In Portuguese such forms as boa b}i from bona through 
*B^il are probably due to the analogy of the masa bom B^i : 
nasals between vowels do not seem to be dropped. On the 
contrary, whenever a nasal is retained, the tendency is to 
give up any distinct nasalizing of the preceding vowel. This 
is the case in French, not only within words, as ia fsmme >]f, 
but also when two words are run together as in son enfatii 
sjl jj>j5 compared with sonj)ere sjj DJJei. 

Vowels tend, of course, to lose their nasality even when 
not followed by a nasal, especially when unstrest. The Old 
Bulg. nasality has been lost in all the living Slavonic lan- 
guages except Polish. 

155. The following are, therefore, the natural stages of 
nasality: 



(I) p 


3jei3 


3^3 


(a) 3n 


3'J93 


3n3 


(3) 3» 


M 


3»'»3 


(4) 3» 


M 


3^3 


(5) 3» 


M 


i'n 



166. But before nasality is dropt, it often considerably 
modifies the quality of the vowel. In the high X^ and h the 
nasality is not very distinct, and there is a tendency to make 
it more audible by enlarging the oral passage. Hence while 
in, im is still li in Portuguese sim^ it has become li in French 
vin. Again, jf etc. have a deeper pitch than the corresponding 
un-nasal voweb, and hence there is a tendency to exaggerate 
the effect by rounding; and when the nasality has been 
removed, the resulting j4 may follow the rising tendencies of 
high vowels, and finally become h. Thus Germanic *gansi 
was borrowed by Old Bulgarian in the form of gqA eij5sX, 
which in Russian has become giuA oAs\ just as *gansi has 
passed into (guws) in Mn£ through ^eqs^s and 0£ gos. 

167. Forward nasalization appears occasionally in Portu-* 
guese, as in mde rr^sis from mdter^ an example which also 
shows that diphthongs may nasalize both their elements, as 



I 



40 HISTORY OF ENGLISH SOUNDS. 

is always the case in Portngaese. In Russian all (back-) 
round Towels are nasalized by a preceding nasal, as in muH 

158. Nasalizing of a preceding cons, is seen in the Swedish 
Aamn from Aavn, lugn ooImi. 

Parasiting. * 

169. Very important is the development of parasite- vowels 
before and after certain consonants, especially the vowellike 
r and /. It appears, however, to be partly due to acoustic 
tendencies. The first stage in parasiting (svarabhakti) is seen 
in such words as E. botoer^ German bauer from older bwr, in 
which the glide to the CD has been exaggerated into an inde- 
pendent 1 or *[. In the affected pronunciation both of London 
and of Berlin this mixed vowel is often developed into a 
full (a). This is no doubt the way in which in French the 
Low German knlf became canif. The quality of the parasite 
is often determined by that of the nearest accented vowel, as 
in Welsh ami 3f3cd, ochr 3<)J(D, OE bur{u)gy *f(*V- ^P § ^4^- 

160. Parasiting implies, of course, a certain difficulty or 
delay in passing from one cons, to another. Hence it rarely 
occurs between two conss. formed in the same place, as 
between I or % and t; here, on the contrary, the tendency 
is towards absorption of any intervening obscure vowel (77). 

lei. E. fear >IX((d) from OE fer shows how parasite-diph- 
thongs begin. Their further development is partly the result 
of divergence, by which w, e9 become ia, ea^ partly of the 
further influence of the vowellike that caused the parasiting. 
Of these influences rounding is the most marked. In OE e 
before r + cons, regularly becomes eo, as in eorpe^ no doubt 
through *e9rfie. The same influence of I is shown in the Tudor 
E. pronunciation of 9alt^ etc., as (sault). In Dutch zout ' salt' 
the I has rounded not only the glide but also the a, and has 
then been itself absorbed, as in E. walk from Tudor (waulk), 
etc. We see the same rounding influence of I in the E. 
pronunciation of children as (tjuldran). This influence of t 
and I seems to be acoustic rather than organic — due to imita- 



SOUND-CHANGE. 4 1 

tion of the deep pitch of these consonants when formed with 
hollowing of the blade of the tongue (104). 

162. The influence of r is, however, generally more in the 
direction of backing and lowering than of rounding, as in 
E. star from Middle E» ^terre. 



Other Influences. 

163. There are, besides, a variety of less important influ- 
ences of consonants. Indeed almost every cons, modifies the 
preceding vowel more or less. Thus in E. the % in jUh has 
not quite the same sound as in hUn. 

164. The most important, perhaps, of these minor influences 
is point modification, by which an (O-position is anticipated 
in the preceding voweL The effect is most marked if the (0 
is inverted. When a vowel has once been modified in this 
way, the (O itself is often dropt. Thus in the Kentish dialect 
Bparroto has become Sd]«vc through *Sd3»(vc)(Dc. In Swedish r 
followed by point conss. and s draws them back to the rim of 
the palatal arch (half-inverted), the r being itself dropt> 
surviving only as a slight modification of the preceding 
vowel : bam Dj^lc, kors Q'jsc. 

166. The general influence of conss. on height and narrow- 
ness is obscure. In Germanic, nasals raise e and \^ i and u 
as in OE 9inga% 9ungen^ while in Danish they widen a pre-> 
ceding » as in finde ^inX, In Danish back conss. have the 
same effect, while in early MnE they narrow a preceding 
i (786). 

The development of i or ^ before initial % + cons., as in 
Spanish escuela, Welsh ys^ol ^ school ' is no doubt phonetic, 
« and e being acoustically allied, because of the highness of 
their pitch ; i is indeed acoustically the i among conss (15). 

There still remain some special influences of cons, on 
cons. 

166. The opening influence of ^ on a following i is seen in 
the Dutch scAip sciItd from sUp, and is parallel to the aspi- 
rating effect of i (140). The later Germanic change of cs 
into as, as in German seeks, is exactly contrary, being probably 



42 HISTORY OF ENGLISH SOUNDS. 

due to striviBg after distinctness and definiteness of arti- 
culation. 

167. Change of place is most frequent in the nasals. The 
change of (ng, nk) into (qg, t)k) is general. In most languages 
d never occurs except before a back stop, a natural result of 
which is that in JQ the superfluous a is often dropt. In Qm 
en in final syllables drops its vowel and follows the place 
of a preceding cons. : sat/en S'3t€U, lieben ooftDr. 

168. One of the most marked changes of place is that of 
kio into jD, as in Greek j)enie, Welsh pump olrD compared with 
Latin quinque, which is apparently against the principle of 
gradual change. The most probable explanation, however, 
is that the 8 was anticipated in the a, first by rounding, 
and then by simultaneous lip stoppage, the back stoppage 
being then dropt as superfluous, and, indeed, almost inaudible : 
03, Qos, Qoio = (a + d)oo, d^o, od, d. 

169. We have already traced the fronting of a and o up to 
the stop-open qo (140). This is the Swedish pronunciation of k 
before front vowels, of iy, and of tj\ as in kind QoIrj*m, tjock 
QoJ-Q*. But there is a natural tendency to shift the com- 
bination forwards towards the more flexible tip of the tongue. 
Accordingly, in South Swedish we find the stop moved 
forward to the blade position, the open element being also 
moved forward, giving dk>»-. The next step is to convert 
ot- into e by keeping the point up instead of lowering it, 
which gives the E. and Italian ck, both of which arise from 
fronted i, as in ckMe from OE ceosan^ cielo from caelum. 
Another change of qo, in a totally diflerent direction, is 
effected by dropping the stop, which is done in Norwegian, 
so that Sw. Qoh^JD becomes oil*. So also Sanskrit gatdm 
o3o3r = Lat. centum. 

170. oe itself, whatever its origin, is liable to further 
changes. If the o becomes pure point, the e naturally 
becomes s. This has been the case in the Dalecarlian dialect 
of Swedish, where qo becomes (ts). So also ch ue became (ts) 
in some of the Old French dialects. In Italian pozzo from 
puteus through *d}do}, the t was probably only slightly 
fronted, so that 00 may have passed almost directly into OS. 



SOUND-CHANGE. 43 

(tj) and (ts), lastly, may drop the first element, giving J — as in 
the present French and Portuguese pronunciation of ch — and «, 
as in the Old Bulgarian «i{^= Latin centum^ where Lithuanian 
has J — 9z\mta9, 

171. The development of the voiced md) is parallel, except 
that dropping of the first element is much commoner, even in 
languages which retain the stop of QO. Thus in Swedish 
and English kind Qof'lHD and gdra (D{«(03, chin and yell (OE 
gellan) are not parallel, although in Dalecarlian Sw. initial 'soft ' 
g becomes dz, parallel to ts from soft k. In Italian, too, soft g 
is (De, parallel to c=oe. So also the later developments (tJ) 
and (ds) are often unparallel, as in Old Bulgarian, where 
iloviku, dogi have vocatives dlovide aeoo3->[»Dej[, boze Dje|). 

172. In most languages there is a tendency to make so>, s\ 
into e, as in E. nation from ME ndsim. 



ACOUSTTO CHANGES. 

178. Acoustic changes may be isolative or combinative. 
Such isolative changes as u to > and (o to ei, which are 
probably, in part at least, imitative, have been noticed already 
(96, 93). The most unmistakeable instances of imitative in- 
fluence are afforded by certain changes between narrow and 
wide vowels. 

174. If we start from a high-wide vowel, such as X, we 
shall find that the nearest vowel in sound is not [, but the 
narrow-mid [, while the nearest in sound to |) is another 
narrow vowel, the low x« This agrees with the pitches of 
these vowels (8), for while C is a whole tone lower than 
X, there is a descent of only a semitone from X to [ ; in fact, 
the series X [ [ *[ X forms a descending chromatic scale. It also 
agrees with the height of the tongue, for the flattening of the 
tongue in X widens the passage more than with X but not 
so much as with [, where the whole body of the tongue is 
lowered. The same relations^ exist not only between the 
front-round, but also between the back-round vowels. The 
unrounded back vowels may be disregarded here. The 



44 HISTOfiY OF ENGLISH SOUNDS, 

following pairs of wide and narrow vowels are, therefore, 
very similar in sound : — 

f } and [e; [e and ^se 

iy and {e; {pandfce 

iu and }o; ^o and jo 
Some phoneticians, sach as Jessen, have even gone so far 
as to maintain that f and [ are one identical sound, which 
may be called indifferently * open i' or * close eJ Cp Cooper's 
vowel-pairs (777). There can be no doubt that the vowels 
in these pairs interchange in language, and that the change 
cannot be explained organically, and is, therefore, imitative. 
The short e in men is () in Southern, but \ in Northern English. 
Open short i in Danish is f, but the parallel 1 has been sup- 
planted by }), as in hundet contrasting with binde^ which has f . 
The lengthened I of OScand. vita^ which is still It in Ice- 
landic has become [t in Swedish veta^ Danish having Ik in 
vide. 

V 

176. Again, we can lower the pitch of I either by rounding 
or retraction, and in the resulting f and I these modifications 
balance one another exactly, so that the two vowels have the 
same pitch, and are very similcur in sound. This gives us the 
following pairs of acoustically similar unrounded mixed and 
front*round vowels of the same height : — 

II and fy; li' and fy 
l^e and {e; \e and {p 

ja and foe; x^ tui<^ {^ 
The present pronunciation of French le with some variety 
of { or { is probably an example of these changes. A still 
clearer one is the change of Arian u into Old Bulgarian y, as 
in 9ynu^ pronounced sin in Russian. As Old Bulg. expresses 
Greek i*=f by v, not by y, in such words as «»w(>= Greek 
muron^ it is tolerably certain that y had the same sound in 
Old Bulg. as in Russian. But it seems certain that y was 
once a round vowel in Russian, for it rounds a preceding 
cons, just like t«(i5i), as in my rpl$. Hence we may assume 
that original 1 became f , as in French, and that this f became 
I by imitation. 

176. Many changes can be accounted for by the stiiving 



SOUND-CHANGE. 45 

after greater audibility. Such are the trilling of r, the change 
of the lip to the lip-teeth >, the exaggeration of the almost 
inaudible breath nasals into nA, etc., in Welsh (105). Others, 
are, partly at least, the result of exaggerating distinctive 
features, as when (low) back vowels are rounded, so as to 
lower their pitch still more. 

Of combinative changes, many appear to be partly organic, 
partly acoustic, such as vowel-harmony (141) and parasiting 
(159). All cases of divergence, whether in diphthongs (67) 
or in consonant groups (105) are mainly acoustic, being the 
result of striving after distinctness. 



EXTERNAL CHANGES. 

177. External changes seem generally to fall under the head 
of analogy, or levelling of distinctions. 

178. Formal analogy is seen in the frequent cases in which 
an originally independent or shifting stress becomes fixed on 
one syllable, as when the shifting accent in Greek pdda^ podSs 
becomes fixed in poim^na, poim^nM^ or when the free Russian 
stress becomes fixed on the last syllable but one in Polish. 
In the Scandinavian languages / between vowels is voiced, 
while s retains its breath sound everywhere; but in E. s 
follows the analogy of the other hisses, and becomes (z) be- 
tween vowels. So also initial s follows the analogy of t?, and 
becomes z in Dutch. It is, however, often difficult to tell 
whether such changes are not, partly at least, organic. 

179. Logical analogy, on the other hand, is entirely inde- 
pendent of organic considerations, often indeed of acoustic 
resemblance as well, being due to similarity of meaning. 
Thus, in MnE the pret. bare has become bore, and in Qm 
the pret. *scAneit has become schnitt because of the analogous 
meaning of the pret. participles borne, geschnitten. This analogy 
is extremely frequent in inflectional and derivative elements, 
as when the OE plur. steorran has become stars in MnE by 
the analogy of the numerous OE plurals in -as. 

180. Another form of logical analogy is the famUiarization 



46 HISTORY OP ENGLISH SOUNDS. 

of foreign words (volksetymologie), as when (uparagus becomes 
^parrow-gras^^ carbunculu9 becomes karfunkel in German. But 
such changes are partly — especially in their beginning — formal, 
being due to the attempt to substitute familiar for unfamiliar 
syllables, for it is not only strange sounds that make a diffi- 
culty, but also strange combinations, whether in sound-groups 
such as initial German U-^ or in syllables. 

181. External influences may be complete, as in 9tar9y or 
partial, as in horCy which is still distinguished from home by 
the n of the latter. They may also be one-sided or mutual, 
the result of a compromise between two forms being called 
a * blending/ 

182. That conscious modification of one of a pair of 
homonyms by which such differentiations (scheideformen) as 
MnE (waind) verb, and (wind) subst. are supposed to be 
obtained, cannot be maintained as anything but a merely 
apparent cause of change. All organic changes are carried 
out without any regard to the logical consequences, as we see 
in such a sound-group as E. (bear), which stands for four 
distinct words, the infin. and archaic pret. of a verb, the name 
of an animal, and the adjective. Of course, if two words 
which would otherwise become identical diverge under special 
influences, the chances of their preservation are increased, as 
when hear and hare were differentiated as heat and hore. If, on 
the other hand, real obscurity results from two words actually 
running together, one of them is simply discarded for a dis- 
tinctive one, as when plough supplanted the verb ear. This 
is also an example of how a language made up of various 
dialects — as all languages are which are spoken over an 
originally diversified linguistic area— can choose the most 
distinctive forms from these different dialects, for plough is 
a Midland and Northern, not a Southern word. Most differ- 
entiations can be explained in this way. Thus hale is the 
Northern, whole the Southern descendant of OE hdL 

188. But although logical considerations cannot alter the 
direction of change, they have a great power of retarding it. 
Every language at any given period is the result of an in- 
cessant struggle between the organic tendency to change, and 



SOUND-CHANGE. 47 

the logical effort to get rid of the resulting ambigaities and 
complexities. If we consider that the consonant-mutations 
of Celtic, the sandhi of Sanskrit, the assimilations of Russian, 
the Germanic umlaut, the Old Bulgarian dropping of final 
consonants, so far from being mere vagaries of Celtic, Sanskrit, 
Russian, Germanic, and Old Bulgarian respectively, are ten- 
dencies common to all speech, we cannot help seeing that 
the unrestrained working of these tendencies through a few 
centuries would make any language utterly unfit for the 
communication of ideas. There are three main results of 
phonetic change against which logic specially revolts : (i) ob- 
scuration of the identity of a word, as when 'head' in Welsh 
is expressed by DfrF, dJJt, :^ or r^[i, according to the ending 
of the preceding word, and when Sanskrit lam D3r= Greek tSn 
appears also in the form of o3^ d]dj d]L, d]i, ojox, ojo), u](Di 
according to the beginning of the next word ; (2) divergence 
of formations from the same root, especially inflections, as when 
in Old Irish we find toibnim 'I drive,' dosennal *they drive,' 
iafneiar * they drove,' toffund * to drive,' all formed in accordance 
with strict phonetic laws from do + 9vand, these manifold di- 
vergencies being mainly due to shifting of stress ; (3) levelling 
of distinctions, mainly through dropping of sounds, of which 
£. supplies many instances, such as the various meanings of 
(beer), the loss of adjective inflection, etc. Logic is not only 
hostile to the confusions that result from sound-change, but 
also to sound-change itself. If language were wholly rational, 
if every idea were represented by one unambiguous word, every 
syllable, every sound of which had a definite logical function, 
the intellectual would have so complete a control of the 
mechanical tendencies of language that sound-change would 
cease altogether. But as language is only partially rational, 
these two tendencies co-exist, the logical element, however, 
predominating, at least in real living, spoken languages — not, 
however, in artificial literary ones. In practice, irregularities 
such as the Olrish toibnim^ etc., are allowed to accumulate 
till they become a strain on the memory, and then the whole 
system is reformed by selecting certain typical forms under 
which all divergencies are levelled, as if Irish were to con- 



48 HISTOET OP ENGLISH SOUNDS. 

jugate toibniniy ^toibnatj ^toibnetar, ^toibnnnd, A striking in- 
stance of such a reaction in favour of grammatical symmetry 
is afforded by the Germanic vowel-gradation (ablaut) in ^inff 
%ang 9ung^ drink drank drunk^ etc. But as internal vowel-change 
obscures the identity of a word, these forms came afterwards 
to be regarded as irregular/ and have accordingly been 
greatly curtailed in favour of the 'regular' conjugation lovedy 
etc., which is generally unaccompanied with internal change. 
It will be observed that grammatical regularity is often directly 
opposed to phonetic regularity: toibnim, dasennat^ etc., are 
phonetically regular, while the levelling Germanic ablaut is 
phonetically irregular — ^to a great extent at least. Hence the 
symmetry and simplicity of the Sanskrit and Gothic vowel- 
system, with its three short vowels a, i, «, is no proof of 
primitiveness, but rather of the contrary. The arresting of 
ambiguity-causing changes is determined by similar practical 
considerations. In polysyllabic languages, such as Greek and 
Old Bulgarian, final consonants could be dropped freely with- 
out making the context unintelligible, but in English any loss 
of final consonants, or, indeed, even so slight a change as that 
of d into ^, etc., would make the language unintelligible. The 
tendency to drop final consonants is as strong in E. as ever 
it was in prehistoric Greek, as anyone may convince himself 
by listening to the listless, slovenly speech of every-day life. 
Every time we €U3k our interlocutors to repeat what they are 
saying, we are really making a logical revolt against final 
consonant-weakening or some other organic change. The first 
Greek, on the other hand, who said gdla instead of gdlai(t) 
was not interrupted with a 'what?,' simply because the word 
was still perfectly intelligible. The extraordinary freedom 
from assimilative influences (sandhi, etc.) which we observe 
in the E. consonant-system is also a result of logical neces- 
sities. This clearness of our consonant-system enables us, on 
the other hand, to weaken our unstrest vowels with impunity, 
while in French the conditions are exactly reversed. We see 
then that every language is forced to resist 9ome phonetic 
tendencies^ while resigning itself more or less completely to 



SOUND-CHANGE. 49 

others. Hence the necessity of comparing different languages 
in ascertaining the general laws of sound-change. 

GENERAL PRINCIPLES. 

184. We have now surveyed the whole field of sound- 
change. We have seen that the organic and acoustic laws 
of change are continually crossed by logical tendencies, as 
when, for the sake of distinctness, the elements of a diphthong 
are diverged, instead of following the organic tendency to 
convergence. 

186. The explanation of the logical and acoustic changes 
is self-evident; not so that of the purely organic. If we 
survey these as a whole, we perceive two principles of 
economy : — 

(a) dropping of superfluous sounds, as when (]gg) becomes 

(9); 

(b) ease of transition from one sound to another, which 
leads to convergence and assimilation, as when (dn) becomes 
(nn). 

186. It is evident that these principles do not help us 
to determine the relative ease of articulation of individual 
sounds, for (r|) in (i|g) is not dropped or modified from the 
desire of easing a difficult articulation, but simply because 
it is superfluous. There seems, indeed, reason to doubt 
whether the inherent ease of an articulation has much to do 
even with isolative change.. As a general rule, all familiar 
sounds seem easy, all foreign ones difficult and harsh. There 
can, however, be little doubt that some articulations, such as 
the trilled point r, do offer some difficulty even to a ver- 
nacular tongue, and that the back ei, which in almost every 
language is substituted for it by individuals, is essentially 
easier, the uvula being simply lifted up by the back of the 
tongue so that it vibrates passively. It is also clear that 
direct isolative changes are from back to front, and from front 
to foreward, and scarcely ever the reverse way. This seems 
to be the result of the superior lightness and flexibility of the 
foreward articulations as compared with those of the heavier 

£ 



50 HISTORY OF ENGLISH SOUNDS. 

root of the tongue. But there are also considerations of dis- 
tinctness. In the first place, the foreward articulations are 
more visible, and therefore more easily learnt by direct in- 
spection ; and secondly, a far greater vaiiety of sounds can be 
produced in the fore part of the mouth : if we were to make 
S, e, u into back sounds, they would all be merged in the one 
sound c. This last consideration is so decisive that we are 
compelled to admit that however probable an organic ten- 
dency from back to foreward may be, there is no absolute 
proof of it. 

187. In many cases we can see nothing but a continual 
fluctuation between two closely allied sounds : we find (%) 
becoming (d) in one language, (d) becoming («) in another ; 
in Danish (S) became (d), and now this (d) has returned to 
nearly its original sound ! So the question is, to some extent, 
one of stability and instability. As regards place, we may 
say that the front consonants are the most unstable, because 
they can be shifted either backwards or forwards, and we find, 
as a matter of fact, that the most unstable consonants are the 
front stops, q and m. <D is saved from place-shifting by its 
vowellike character. The vowel 3 is very unstable, because 
it can be modified in the direction either of (o) or of (e). 
Long vowels are more unstable than short, because the 
longer the sound, the more temptation there is to modify it. 
The most stable vowels ought therefore to be the short fronts. 
We find accordingly that original Ar. short i and e have been 
preserved up to the present day in such words as toil and 
seven. Compound sounds, such as the rounded vowels, are 
of course unstable, as shown in the development of short u 
into Swedish 1^, E. ], and in French f from Lat. it. 



OEIGIN OF SPEECH-SOUNDS. 

188. It used to be generally assumed that primitive speech 
had a very limited range of sounds ; but a little consideration 
will show that the opposite must have been the case. Lan- 



ORIGIN OF SPEECH-SOUNDS. 5 1 

guage proper, which implies sound-groups (words) symbolizing 
ideas, and capable of being combined into sentences as freely 
as ideas are combined into thoughts, was preceded by a 
period of mixed gesture and imitation. Every object and 
phenomenon associated in nature with an imitable sound 
would naturally be named by an imitation of that sound : 
Qlai — or some such sound-group — meant ^ cuckoo ' from the 
beginning. The power of imitation was enormously developed 
through its use by hunters in decoying wild animals, where, 
of course, the best imitation would secure the best results. 
But gesture also helped to develop the power of forming 
sounds, while at the same time helping to lay the foundation 
of language proper. When men first expressed the idea of 
' teeth,' ' eat,' or ' bite,' it was by pointing to their teeth. If 
the interlocutor's back was turned, a cry for attention was 
necessary, which would naturally assume the form of the 
openest and clearest vowel (a). Sympathetic lingual gesture 
would then accompany the hand-gesture, which latter would 
then be dropped as superfluous, so that (ada) or, more emphati- 
cally, (ata) would mean ' teeth ' or ' tooth ' and ' bite ' or ' eat/ 
these different meanings being only gradually differentiated. 
We see that the primitive uninflected words or ' roots ' of 
language were probably dissyllabic. So also the ideas of 
* wind ' and ' breath ' were expressed by o + vowels, which is 
both an imitation of the sound of the wind and is at the same 
time one of the results of the action of breathing itself, ' blow- 
ing' being also expressed by 0. Now neitiier o nor form 
part of the original Arian sound-system, as known to us by 
historical evidence. Not only isolated sounds like o were 
eliminated, but also whole classes of sounds. Primitive man 
must have expressed 'drinking' by an inbreathed c<, and 
probably he expressed sensual enjoyment generally, as some 
of us still do, by an inbreathed voiceless ^--o^. These incon-> 
venient inbreathers seem to have been eliminated everywhere 
in language, but the nearly-related suction-stops or * clicks ' 
still survive in many primitive languages, as in the South 
African Bushman and Hottentot, and in some Califomian 
languages. These clicks were no doubt originally (as pointed 

£ 2 



52 HISTOfiT OP ENGLISH SOUNDS. 

out to me by Mr. J. Maxshall, junr.) food-cries. Another 
dasd of soandB which have been eliminated in most languages 
is that of the throat-consonants or ' true gutturals/ which 
still survive in Arabic, and also seem to have existed in 
parent Arian — at any rate, in Sanskrit. But the Sanskrit 
' sonant k ' may be a new formation, like the glottal stop in 
Danish. Clicks still survive as interjections in English. 



OEIGIN OF DIALECTS. 

189. Language originates spontaneously in the individual, 
for the imitative and symbolic instinct is inherent in all 
intelligent beings, whether men or animals ; but, like that of 
poetry and the arts, its development is social Where there 
is free and uniform intercourse between all the members of 
a community the language will be uniform — that is, uniform 
in the sense of not splitting into dialects. Of course, every 
family, and every individual, will have their own peculiarities 
of speech, but there will be no local concentration of these 
peculiarities. When the community is too large to permit 
of uniform communication throughout it, dialects begin. If 
we suppose a large plain covered with villages of equal size 
and independence at equal distances, each village communi- 
cating directly only with its immediate neighbours, there 
will in a few generations be a distinctly different dialect in 
each village, and in course of time the dialects of the most 
northern, southern, eastern and western villages wUl become 
mutually unintelligible to one another and to that of the 
central village. But there wiU be no lines of division : the 
dialects will shade insensibly into one another ; the dialect of 
a village halfway between the most northern and the central 
village will partake so equally of the characteristics of the 
northern and central dialects that it will be impossible to 
assign it to either. 

100. This overlapping of dialects — which always happens 
when there is no definite barrier — is due also to the fact that 



ORIGIN OF DIALECTS. 53 

the separate changes which constitute diffei*ence of dialect or 
language do not follow the same boundary-lines, but cross 
one another to any extent. Thus in OFrench the distinction 
between the ' Central French ' or Parisian and the Norman 
dialect is generally fairly definite, but we find South Norman 
agreeing with its neighbour Parisian in changing Lat. c into 
cA (tj) before a, as in cAier=Ls,t. cdrum against the North 
Norman her. This particular sound-change has, then, chosen 
aniMtrea of its own, regardless of the areas of the other 
changes which separate South as well as North Norman from 
Parisian. 

lOL But if such a territory is intersected by a range of 
mountains, a broad river, or any other obstacle to communi- 
cation, running, say, east and west, then there will be a 
corresponding line of linguistic division : all the dialects north 
of the barrier will form a group with features in common 
distinct from those which unite the southern group of dialects ; 
if the barrier is strong enough, the two nearest villages north 
and south of it will in time come to speak mutually unin- 
telligible languages. Even the most trifling barrier — a narrow 
brook or strip of sandy heath — will be enough to mark off 
two groups of dialects. 

192. Complete territorial separation through emigration is 
a self-evident cause of dialectal divergence ; but in such 
cases there is always the possibility of the divergence having 
begun before the complete separation. 

193. There are other factors which disturb the ideally 
uniform development of dialects. In real life, certain villages 
would be sure to gain some kind of ascendancy over those 
nearest it, and thus one or more centres of dialectal influence 
would be established ; till at last, if centralisation were strong 
enough, one dialect would be used as a means of expression 
all over the territory, as is now the case in England. If 
communication and education were made perfect, the standard 
dialect would entirely supplant the other dialects, and absolute 
uniformity of language would prevail. 

194. In this way political development also tends to cause 
definite lines of division, for each linguistic centre swallows 



54 HISTORY OF ENGLISH SOUNDS. 

up the dialects nearest to it, till it comes in conflict with 
another centre, the Ime of division generally, though not 
necessarily, coinciding with some natural boundary. Hence, 
if we compare two standard languages of the same family, such 
as Dutch and German, we are struck by their fundamental 
difference, and have no hesitation in calling one Low, the 
other High German. But if we compare the dialects of the 
two languages, we shall find them shading off into one another 
by insensible degrees, there being many 'Middle Germasi' 
dialects which carry out the change of t into OS, as in zeit, 
but leave initial j) in its unaltered. Low German stage, as in 
jmndy the present standard German being itself a dialect 
intermediate between High and Low. 

196. It need hardly be said that the standard and the local 
dialects influence one another strongly. Standard E., which 
is mainly East Midland, has taken words and forms from 
almost every other dialect ; vat^ for instance, is Southern, 
Aale (^loiole) Northern. 

196. Not only dialects influence one another, but also 
languages, even if they belong to totally distinct families. 
Thus Finnish is full of archaic Germanic and Lithuanian 
words, Persian is mixed with Arabic, and so on. Even 
sounds are borrowed. Thus the southern Bantu languages 
in Africa have borrowed the clicks from the Hottentots: 
Zulu has them, but they are wanting even in Bechuana. So 
also the peculiar ' choke-stops ' of Armenian (ca, etc.) have 
been borrowed from the non-Arian languages of the Caucasus. 
Sanskrit, again, got its inverteds from the Dravidian languages 
of the South of Lidia. English and Welsh too, with their 
()?) and ("8) and their (w), have much in common. There is 
no limit to the mixture of languages in sounds, inflections, 
and syntax as well as in vocabulary. But the influence is 
never equal on both sides. Finnish has borrowed largely 
from Germanic, but there are very few common Germanic 
words of Finnish origin. So also the proportion of English 
words in spoken Welsh is about the same as that of French 
words in Chaucerian English, but there are very few Welsh 
words in English. Li fact a very intimate mixture of two 



ORIGIN OP DIALECTS. 55 

languages is always a prelude to the complete extinction of 
the weaker one, and this is why few, if any, of these thoroughly 
mixed languages become permanently fixed. 

197. Dialects are not only local, but social, as in the dis- 
tinction of polite and vulgar speech, vulgar speech being 
generally ahead in its development, as in the Cockney and 
dialectal dropping of (h) in E. There is also the important 
distinction of the literary and colloquial dialect, the former 
being mainly a written dialect, consisting of a mixture of 
living colloquialisms with the colloquialisms of earlier stages 
of the language, as when in poetry we use the fossilized 
colloquialism tAou Aast side by side with the living collo- 
quialism you have. Of course, when the divergence amounts 
to unintelligibility, as when an Italian writes Latin, we have 
two distinct languages, a dead and a living, the latter being 
still liable to be influenced by the former, these influences 
spreading even to the vulgar dialect. Such languages as 
Latin and Sanskrit, when written and spoken by modem 
scholars and pundits, are commonly stigmatised as ' artificial' ; 
but the artificiality is not in the languages themselves, but in 
the means by which they are preserved — in the case of Latin 
by written symbols, in that of Sanskrit by an uninterrupted 
oral tradition. This preservation of a dead language is, 
however, never perfect. Li the first place, the process of 
fixing is always at first tentative and inconsbtent — even 
Sanskrit embodying colloquial Prakrit forms — ^and secondly, 
it is impossible to fix the pronunciation, as is again clearly 
shown in the present pronunciation of Sanskrit, in which 
some of the sounds, such as bA and f, are confounded, and 
others much modified, partly by the influence of the living 
Gaunan languages, but apparently also by natural develop- 
ment of Sanskrit itself after it had ceased tp be a colloquial 
language. 

188. External circumstances not only have an influence on 
the development of dialects, but they also directly modify 
the sounds of a language. Climate has some, though a very 
slight influence. Li cold countries there is less disposition to 
open the mouth widely. Hence that tendency to make a into 



56 HISTOET OF ENGLISH SOUNDS. 

which is almost umversal in the modem Germanic languages, 
but is quite absent from the Bomance languages. The dis- 
position of the speakers may also influence their pronunciation. 
The habit of speaking with a constant smile or grin unrounds 
the vowels, as in the Cockney (nau) = «^. The refinement 
and effeminacy of large cities untrills the r. Even the 
caprices of fashion may have their effect, as is shown in the 
lisping pronunciation of those savages who knock out their 
front teeth. 

199. Not only every language and dialect, but every period 
of a language has its own laws of change, and its own sound- 
system, which includes only a few of the possible sounds and 
their combinations. There is nothing to prevent two closely 
allied languages or two periods of the same language from 
following opposite tendencies. A group of languages like the 
Bomance may agree in a dislike to harsh consonant-groups^ 
but this does not prevent Portuguese from consistently drop- 
ping its weak e in such a word as vistes ' ye saw,' which is 
now colloquially ^Iziozu A language may unround all its 
(y)'s, etc., into (i)'s in one generation, while its (u) is moving 
in the direction of (y), so that the front-round vowels again 
come to form part of its vowel-system. 

200. This last case also exemplifies the perpetual loss and 
re-development of a sound. As a general rule, it is the most 
distinctive sounds which are most quickly restored. There 
may be periods in any language in which such vowels as a, » 
or u are eliminated by various changes. Thus in Early Mod. 
E. there was a period when the I of OE win had become (ei), 
while the nearest approach to (ii) was the [f^=OE tcen^ but 
(ii) was soon restored by further raising of this e. 

201. Languages which are very rich in sounds, such as 
Sanskrit and Kussian, generally owe it to assimilative in- 
fluence. The difference between a poor and a rich sound- 
system is merely that the former utters the elements of such 
a group as 'JO)] successively, while the latter utters the first 
two simultaneously — Tx] or l3, the former -class of languages 
being generally more harmonious than the latter, which often 
have something 'sloppy* about them. We find, accordingly, 



OKIGIN OF DIALECTS. 57 

that many of the Sanskrit sounds, such as oi and (D(, occur only 
in special sandhi-combinations. After what has been said 
about the richness of primitive sound-systems (i88), it need 
hardly be repeated that extreme simplicity is no proof of 
the primitiveness of a sound-system, being, as often as not, 
the result of levelling, as in Qothic, where e and o were 
levelled under i and u respectively, or being only apparent — 
the result of a defective alphabet, as in the Old Persian of the 
cuneiform inscriptions. Languages spoken over a diversified 
linguistic area tend to simplify their sound-systems, as may 
be seen by comparing German and Italian with any of their 
dialects, most of which show complex sound-systems. 

202. No language has an absolutely symmetrical sound- 
system, because every sound-system is the result partly of 
organic, partly of logical influences. The organic tendency is 
towards analogy and symmetry. Such organic changes as 
the unrounding of front vowels are generally carried out con- 
sistently: if we hear a German say (giita) instead of ffiite, 
we expect him to say (Jeen) instead of scAon. There 
is also an organic tendency to carry out a uniform basis 
of articulation. Thus the English tendency is to flatten and 
broaden the tongue, which makes the vowels wide, and to 
hollow the fore part of it in forming such conss. as / and ^, 
which tends to draw away the tongue from the teeth. If this 
tendency is exaggerated, it results in a general back-modi- 
fication, which would end in nm.lriTig our concave / into a 
Russian 8. In E. there is also a tendency to keep the mouth 
half shut, which is partly due to the climate (198), and is the 
first step in the direction of rounding. A Frenchman, on the 
contrary, articulates with a convex tongue, either against the 
teeth, or as near them as possible, and opens his mouth 
widely. But the carrying out of a uniform basis of articu- 
lation would often lead to the loss of distinctive sounds. 
Thus the dentality of E. ^ is quite inconsistent with the 
general character of its sound-system, but the conversion of ^ 
into s or inner t has up to the present been successfully re- 
sisted by the logical principle of distinctness. But even 
without logical influences we find violations of the basis of 



58 HISTORY OF ENGLISH SOUNDS. 

articulation. Thus in Portuguese t and d are interdental, 
but n is the E. i^^ and I is cd-k, a sound which would seem to 
be totally opposed to Bomance tendencies. 

203. Sound-systems are further characterised by their re- 
lation to the three main modifying factors : assimilation, posi- 
tion, and stress. We must distinguish accordingly : 

(a) assimilation-iDAnence. Does the language allow sounds 
in succession to modify one another, as in Russian, or leave 
them unmodified, as in E. ? 

(6) poHtion-iJifLxience. Are the sounds of the language liable 
to change in certain positions? Has the language 'end- 
laws'? Does it, for instance, throw off all final conss., or 
allow only certain conss. to come at the end of a word ? 

(c) stress'infbience. Does the language modify its sounds 
(especially its vowels) when unstrest, as in E., or has stress 
little or no influence on sound-change, as in French ? 

204. The question now arises. How far can we predict the 
direction of change in a given language ? This will depend 
on the nature of the sound, and how far it has advanced in a 
certain direction of change. In the case of such a vowel as 3) 
all we can say is that if it changes, it will be either in the 
direction of j or of x- But if it has already become J*-* we 
may predict with some confidence that it will become x- So 
also of Q ^e may predict not only that it is very likely to 
change, but that it is almost certain to develop into uz. 
But of the less advanced Q^ it would be impossible to predict 
whether it will advance to Q or return to a. It would, of 
course, be impossible to predict such a phenomenon as the 
Germanic vowel-mutation in a language where the vowels 
had not begun to influence the preceding conss. 

206. Hence, when we see such a phenomenon as vowel- 
mutation developing in all the Germanic languages after their 
separation, we are bound to assume that the change was 
initiated before their separation — that in parent Germanic 
the front vowels had begun to modify preceding conss. So 
also when we find Arian k developing into two sounds inde- 
pendently in Sanskrit and Slavonic without any assimilative 
cause in either language, we are forced to assume that in 



SOUND EEPEESENTATION. 59 

parent Arian k had already separated into two such sounds 
as a^ and Q^. Of course, such a change as that of k into (tj) 
before an i in two separate languages proves nothing, for this 
is the only direction of change possible. It is in practice 
often difficult to decide what weight to give to parallelism of 
change, for languages in a similar stage of development often 
show very striking coincidences which can be proved to be 
quite independent developments, as we see in comparing the 
Romance with the Gaurian languages. 

206. When we find a high- vowel such as (ii) diphthonged 
into (ai), we naturally expect to find a parallel change of the 
other high vowels — we expect to find (uu) becoming (au). 
In such a case as this we are not likely to be mistaken. But 
it must be remembered that two such vowels as (ii) and (uu) 
have nothing whatever in common except their height, and 
that the natural tendency to diphthonging may in the case of 
(uu) be counteracted in some unforeseen way by its rounding 
or back position, so that its diphthonging may either lag 
behind that of (ii), or never take place at all. We have a 
clear instance of this want of symmetry in the MnE levelling 
of ee and ea under (ii) in see^ sea, while the earlier distinction 
between oo and (o)a in moon, moan is still kept up. 



SOUND REPRESENTATION. 

ORIGIN OF WRITING. 

207. Wherever we can trace the history of sound-writing, 
or writing proper — ^the art of representing speech-sounds by 
graphic symbols — we shall find that it was never the result 
of immediate invention, but was evolved by slow degrees 
from the more primitive art of picture-writing with hiero- 
glyphs, whose form more or less directly suggests the idea to 
be expressed, without reference to its sound, as when the sun 
is represented by a circla The first step towards sound- 
writing would be— supposing the language to be written were 
English — to use the cii'cle as the symbol not only of (san)= 



6o HISTOET OF ENGLISH SOUNDS. 

' sol ' but also of (san)=' filius/ and then of the syllable (san) 
or (sa), until finally it came to denote the single sound (s), or 
(s) followed by any vowel. 

208. Such is the origin of the Latin alphabet. It was 
originally an adaptation of one of the Greek alphabets, which 
in their turn were an adaptation of the Phenician alphabet. 
The Phenician alphabet itself was a selection from the 
numerous symbols of the Hieratic writing of the Egyptians, 
which was a compromise between sound- writing and picture- 
writing, evolved by the exigencies of practical life out of the 
older purely hieroglyphic system. 

LAWS OF FORM-CHANGE 

209. The laws of form-change in writing — whether hiero- 
glyphic or phonetic — bear a striking analogy to those of 
sound-change : change is always going on, it is^gradual, and 
it follows definite laws. 

210. Form-change is always going on, because it is impos- 
sible for the human hand to repeat indefinitely the same 
movement without altering its direction and length. Hand- 
writing varies not only from generation to generation, and 
between individuals of the same generation, but also in the 
individual himself, according to speed and care of writing, etc. 

211. Form-changes are partly determined by the nature of 
the material written on and the instrument written with. 
Thus letters cut on stone or wood will be angular and detached, 
while writing with a pen will tend to roundness and joining 
— in short to cursiveness, — writing with a style on wax tablets 
will have a different character from writing with a nibbed 
pen on vellum or paper, and so on. 

212. The most elementary change is one which we make 
unconsciously whenever we write ; a variation in the relative 
lengths either of the strokes of which a letter is composed, or 
of the letters themselves. We see the former change in the 
development of h out of H, the latter in that of j out of i, and 
both together in 1 out of L. 

213. In all cursive writing there is a tendency to round off 



SOUND EBPEESBNTATION. 6 1 

angles, in order to avoid the sudden check and consequent 
waste of force and time caused by an angle, as we see in 
comparing E with e. In a stronger form this tendency leads 
to slurring or degradation, which is generaUy accompanied 
with shortening, as in the second element of r compared with 
B. The tendency of degradation is, of course, to reduce 
originally distinct letters to one form, as we see in the con- 
fusion of y and y in y , etc. Of course, if any element of a 
letter is superfluous for purposes of distinction, there is a 
tendency to drop it altogether, as in b from B, where the 
upper loop of the latter has been discarded. The opposite 
phenomenon of exaggeration of an originally subordinate ele- 
ment of a letter, which is at the same time lengthened, is seen 
in the development of the sidenstroke of a and q into the 
lower circle of g and the upright stem of q respectively, and 
very strikingly in the development of the Black Letter or 
Gothic alphabet, in which originally merely accidental and 
ornamental tags have been exaggerated so as to obscure the 
original elements of the letters. These changes are, of course, 
due partly to the organic tendency to variation, but also to 
the stiiving after distinctness. While there is a general ten- 
dency to round off angles, as in c from <, there is a tendency 
not only to keep acute angles, as in our w compared with u, 
but also to turn sharp curves into angles, as in the develop- 
ment of f out of s. 

214. In writing with a nibbed pen the down strokes 
are thick, the upstrokes are thin — a peculiarity which still 
attests the origin of our printing letters from quill- or reed- 
written ones. Hence the tendency to employ the thicker 
and distincter down-strokes as much as possible. It is 
easiest to thicken a down-stroke when it is more or less per- 
pendicular, and as variations of slope are inconvenient in 
many other ways, all but perpendicular down-strokes are 
eliminated as much as possible, or oblique strokes are made 
upright, as in q from q. Oblique strokes are often got rid of 
even at the cost of an angle or break, as in d from d. In y, 
X, and some others, the slopes were kept for the sake of sym- 
metry of form, and distinctiveness. 



62 HISTOaT OF ENGLISH SOUNDS. 

216. The above are isolative changes. But there are also 
combinative ones. In writing, the instinct which rebels 
against angles tends also to eliminate breaks as much as 
possible : in all swift and easy writing the letters of a word 
are not only formed individually without breaks, but the 
whole word is written, as far as possible, without lifting up 
the pen. It may happen, as in the case of our script x, that a 
letter may have a break in it, and yet be joined without a 
break to the preceding and following letter. This peculiarity 
was strongly developed in the Old Roman cursive hands, as is 
seen in the Ravenna papyri, the result being that the shapes 
of individual letters varied according to their position and 
combinations. We see the results of this system in the 
Arabic alphabet, where many letters have three different 
forms — initial, medial, and final. Even in the modern Latin 
alphabet we have — or had, till lately — the distinction of 
initial and final s and medial f. 

216. The final result of unchecked organic changes would 
be to make writing unintelligible. This actually happened in 
the case of the Arabic script. The difficulty was met by the 
adoption of diacritics : the letters which had run together were 
differentiated by the addition of dots, as many as three being 
sometimes placed on one letter. So also in the Middle- Age 
Latin alphabet ni had become confused with m and so on, so 
that the i had to be marked with a diacritic — a clumsy device 
which we are still forced to keep up. 

217. But the logical reaction generally begins long before 
cursive writing has reached the Arabic stage. The first step 
is to detach the letters, selecting from the various cursive 
forms those which are the simplest and most compact— in- 
volving fewest breaks — and the most distinctive. A good 
specimen of such a detached cursive is afforded by the im- 
perial Chancery hand of the Romans. The reaction against 
slurring leads to detaching the strokes even of separate letters. 
Thus we find the top stroke of ; from a, which was originally 
an exaggerated flattening of the top curve, completely de- 
tached in the oldest Roman cursive writing, and so with many 
other letters. One of the most effective means of securing 



SOUND EEPEESENTATION. 63 

simplicity and distinctiyeness, is by utilizing projection above 
and below the line, which developed itself spontaneously in 
the Roman capital writing, and after much fluctuation settled 
down into the usage of our present minuscule or lower-case 
alphabet, in which, for instance, i j 1 represent distinctions 
what were once almost entirely dependant on projection. 

ALPHABETS, 

218. The angular and detached letters of the Roman lapidary 
alphabet were, however, modified differently for different 
purposes. The old alphabet was used for writing books long 
after a fully developed cursive had come into use for the 
ordinary purposes of life, this cursive itself being nothing but 
a degradation of the book alphabet. In the ' uncial ' alphabet 
A, D, E, M are rounded off in the direction of a, d, e, m, and 
certain letters project above and below the line. The cursive 
writing itself split up into a variety of forms, as in the alpha- 
bet of the wax tablets, the Ravenna papyri, and the detached 
* half-cursive ' Chancery hand. About the fourth century all 
these alphabets existed side by side — as they still do in such 
forms as A a a — and modified each other in various ways. 
A special development of a very old Roman cursive — or rather 
of a degraded capital writing — artificially modified and sys- 
tematised, was the Roman shorthand — the ' Tironian Notes.' 
The chief infiuence of the Tironian notes was on the Middle- 
Age system of contractions, which, again, has in some cases 
permanently influenced the alphabets of modem Europe, the 
Spanish tilde in ano^ for instance, being nothing but the old 
^^-contraction (~), itself probably a degraded M written over 
the line. But the history of the later alphabets is, in the 
main, one of an incessant action and reaction of the detached 
and formal book hands and the cursives on one another, which 
latter were only exceptionally employed in writing books. 

219. When the Roman empire broke up, separate national 
handwritings sprang up in the different provinces in the same 
way as Latin split up into separate languages. A very 
Aarked variety of minuscule was developed among the 



64 HISTOBY OP ENGLISH SOUNDS. 

chriBtianized Celts of Britain, being mainly a compromise 
between uncial and cursive. This alphabet, which is still 
preserved almost unchanged in Ireland, was adopted by the 
Anglo-Saxons, who afterwards adopted "p and p (w) from their 
own Runic alphabet — at first in their original angular forms 
— ^instead of tA and u{ii). They also modified c^ into "S to ex- 
press the sound w, which was probably suggested by the use 
of crossed d (as of other crossed letters) in contractions. 

220. By the time of Alfred the English hand had developed 
a character of its own, the uncial writing having been 
abandoned in favour of the minuscule, from which — at least 
in its book form — many of the older cursive elements were 
eliminated. 

221. The chief subsequent changes were in the tags with 
which the strokes were generally finished off in British 
writing. After about 950 there is a general tendency to 
curve inwards the lower ends of upright strokes in such 
letters as t, n, m, A. About 1050 the ends of low stems are 
curved outwards in such letters as p, f, ]?, while p retained its 
older straight stem. Sometimes these low stems were finished 
off with a cross-stroke or * serif,* as in our printing letters. 
Earlier in the century they began to wave and lengthen the 
top tags of i, n, Ay etc. y occurs dotted in the very oldest 
writings, but the dot was afterwards generally dropped, and 
not restored till about 1000. This, and other changes, were 
partly due to the influence of the French hand, which towards 
1000 began to be generally used in writing Latin. In the 
earlier charters the Latin and English portions are all in the 
British hand, but after 1000 the Latin is in the French, the 
English portions (boundaries, etc.) in the national hand. 

222. This French hand — ^the * Caroline minuscule * — ^was 
developed in France at the beginning of the ninth century by 
a reform of the earlier Merovingian cursive. It is practically 
almost identical with our present Roman lower-case printing 
letters, which were modelled on it. It dots the y, leaving the i 
undotted, and prefers f to «. The stems of the letters are only 
slightly tagged. Its characteristic letters, as compared with 
the English hand, are r,/. y. The upright d and the high i 



SOUND BEPEBSENTATION. 65 

occur in the older English writing, but in Alfred's time they 
had been generally supplanted by d and f , so that their re- 
appearance in Latin writings of this period must be ascribed 
to French influence. 

223. In the first hand of the Peterborough Chronicle, which 
ends at 1124, « ^^^ ^ still retain their English forms, though 
the French d is occasionally used. The high f appears beside f 
not only in this Chronicle, but also in other E. mss even of the 
first half of the nth century. After 11 24, the Peterborough 
Chronicle is written in a variety of hands down to 1154, and 
in this portion the French forms of/, g etc appear for the 
first time in English words, side by side with the British 
forms. Here also occurs the French w^ formed by interlacing 
two v's, but only in French names. 

224. Henceforth writing in England follows the general 
European development. Exaggeration of the tags and stem- 
bending increase, and in the course of the 14th century 
the letters become more and more angular, resulting in the 
crabbed and interlaced forms of the Gothic or Black Letter 
and German alphabets. Then the Humanists restored the 
minuscule of the 12th century. Both types of writing — 
the Latin and Gothic — ^were finally fixed by the invention 
of printing. The influence of the Middle- Age cursives is 
shown in our Italic alphabet. We still keep the old Soman 
capitals unchanged, but only for special purposes of ornament 
and distinction. 

NEW LETTERS. 

226. Every alphabet is liable to the demand for new sym- 
bols either through sound-change in the language which is 
written in it, or through its application to some other language. 
If the change of any sound is carried out regularly in a 
language, the symbol is generally kept also, however much 
the sound may have altered, as we see in French «=f, Italian 
g before e^mt etc. If, however, a sound splits up into two 
different ones with a corresponding difference of meaning, as 
in German grUe^ ^'^ solder guoto^ guoti^ the want of a new 

F 



66 HISTORY OF ENGLISH SOUNDS. 

symbol makes itself felt. Again, in adapting such an alpha- 
bet as the Roman to a new language, the letters will be 
assigned to their nearest equivalents, minute differences being 
disregarded, as when Latin / was used to denote > in Old 
Irish. Often, however, new distinctions have to be made, as 
between I and o in Welsh, or totally new sounds have to be 
symbolized. This is effected in various ways : 

(a) By assigning new values to superfluous letters, as when 
the Greeks made the Phenician o into a vowel-symbol, there 
being no Greek sound answering to the throat-consonant • it 
stood for in Phenician. In this case the change of value, 
though considerable, is by no means arbitrary. Even the 
change by which £, originally the aspirate Ae, was made into 
a front vowel, and the later one by which h, originally the 
throat A'etA «f , came to represent first A and then ^, can be ex- 
plained by the names of these letters, both of which b^in 
with e modified by a mere breath-glide, or what would easily 
be weakened into it. No doubt there may be cases of arbi- 
trary assignments of values, but they are certainly rare. 

(b) By utilizing originally unmeaning variations. Thus, up 
to the i6th century t;wafl simply another way of writing «,and 
J of t: in the 15th century v andy were ornamental varieties 
which were especially used at the beginning of words, and so 
naturally came to be regarded as consonant symbols. So 
also the French 9 is only a variety of a descending z. In 
Olcelandic consonant capitals were utilized as double letters, 
as in maJ\ra:!:= manna, 

(c) By digraphs, such as the tA, ps with which the Komans 
transcribed the Greek ^, </r. Both of these, however, were 
compound sounds o^, ds, so the digraphs are really expansions 
of contractions. But when the Romans expressed Greek initial 
f by rA in rAetor etc, they were using two letters to express 
one simple sound, the A being here a breath-modifier, as if we 
were to express o by (o: in Visible Speech. Of course, when 
tA and pA in Latin became simple u and o, A came to be re- 
garded as an open-modifier. A afterwards came to be a general, 
abnost arbitrary, modifier, to show not only opening and un- 
voicing, but also fronting, as in Provengal and Portuguese lA 



SOUND EEPEE8ENTATI0N. 67 

=00, vowel-length, as in German ohne and E. ah — ^a usage 
which was already developed in Umbrian and Oscan — while 
in Italian gh it was added to show that g kept its original back 
articulation. Doublings are a special form of digraphs. In 
vowels it is a common method of indicating length, as also 
with conss. Some languages which have no double conss. 
use cons.-doubling as a ' strengthener ' or arbitrary modifier. 
Thus in Spanish //=oa, in Weldi=tt, where also rf</=w, f=>,f 
keeping its old British value of >. Greek gg=^JQ is an ex- 
ample of what may be called a compound doubling. Trigraphs 
also occur, as in Gm scA^ Swedish siJziz'K hK, 

(d) By Ugaturee, such as a and a=^ae^ oe^ which in Latin 
were originally diphthongs Jxt, }i!<r, but were afterwards 
simplified to x* sud \k resp. Our w is a consonant-ligature, 
which preserves an extinct form of the vowel u, 

(e) By diaoritios. One way in which diacritics may be de- 
veloped is by writing one letter above another, which was a 
natural device to save space, especially at the end of a line, 
and would easily be utilized phonetically, as in the Geiman t^, 
originally A, where the « is a front-modifier. So also in 
Swedish j=}0 the is a rounder. As we see, such an over- 
written letter soon gets degraded into mere dots or strokes. 
Special contraction-marks were also utilized as diacritics, as 
we see in the Spanish ^ and 0£ 9. Another way in which 
diacritics develop is by degradation of a ligature-letter, as in 
f from a^ where the tail is a degraded a. 

COREESPONDENCB OF SOUND AND SYMBOL. 

226. All writing which has once emerged from the hiero- 
glyphic stage is at first purely phonetic, as far as its defective 
means will allow. But as the association between sound 
and symbol is almost entirely arbitrary, there is always a 
tendency for the symbol to lag behind the changes of the 
sound. 

227. One result of this is the retention of enpexflnouB sym- 
bols, as when we write q instead of £? or ^ in the combination 
qu, this q having originally represented the Semitic inner Q4. 

F % 



68 HISTORY OF ENGLISH SOUNDS. 

The worst form of superfluity is writing ' silent ' letters, as in 
the E. know. 

228. The opposite of superfluity is ambiguity, by which one 
symbol has to represent more than one sound. To some 
extent, this defect is inherent in all sound-notation : even in 
Visible Speech we often omit the minuter glide-symbols etc, 
and in speaking of a practical alphabet we should hardly 
characterize it as unphonetic because it neglected — ^as most of 
them do — to mark even such necessary elements as vowel- 
quantity and stress. If an orthography makes a consistently 
phonetic use of the materials it has : if it restricts every indi* 
vidual symbol to one distinctive sound (which may include 
slight varieties, such as f , f -r in E. jt?%), and does not continue 
to write silent letters, we call it * phonetic' If, for instance, 
in E. the vowels in Uy see, set, say, were invariably expressed 
by i, a, e, ee we should say that E. spelling was, so far, phonetic, 
even if we admitted that the long vowels were really diph- 
thongs. If we found these vowels written respectively i, ee 
e, ai as invariably as on the other system, we should say that 
English wa& ' half-phonetic,' or phonetic on an unphonetic 
basis, for it is evidently unphonetic and irrational to make 
ee the long of t. But when we find such a vowel as 
that in see expressed also by e, ea, i, we must call English 
spelling simply unphonetic. It would be a rhetorical exag- 
geration to call it wholly unphonetic as long as such a symbol 
as ee, together with many of the consonants, retains its present 
uniform value. 

229. We see, then, that unphoneticness is mainly the result 
of the retention of originally phonetic spellings after they 
have become unphonetic through sound-change. It is^ there- 
fore, the result of tradition. Where there is no traditional 
spelling handed down, as when such a language as Old English 
was first written in Latin letters, spelling can hardly help 
being phonetic; where, on the other hand, there is a large 
literature, and, perhaps, a class of professional scribes, the 
influence of the traditional orthography become stronger and 
stronger, till, at last, the invention of printing and the growth 
of the newspaper press make changes of spelling as incon- 



SOUND EBPEESENTATION/ 69 

venient as they were formerly easy. The ideal of a printer's 
orthography is one which is absolutely uniform over the 
whole territory of the language, and absolutely unchange- 
able. Such an orthography as that of the present English 
is, consequently, one in which there is no longer any living 
correspondence between sound and symbol — it is, in intention 
at least, wholly unphonetic : it is preserved by graphic, not 
phonetic, tradition. 

280. But unphoneticness has its practical limits. A purely 
hieroglyphic writing, though cumbrous, would not overtax 
the average intelligence, but an absolutely unphonetic de- 
gradation of an originally phonetic system — one in which the 
separate letters had become phonetically unmeaning — could 
not be mastered even by the most retentive memory. Hence 
a phonetic reaction becomes inevitable sooner or later. In the 
early Middle Ages, when the multiplicity of dialects and the 
fewness of books made a uniform and fixed orthography im- 
possible, the spelling was periodically readjusted in accordance 
with the changes of pronunciation. Thus, when in German 
Au8 had developed into the fully diphthongic (haus) they wrote 
it Aaus. This was easy enough as long as the phonetic tradition 
of the values of the Roman letters was kept up, and as long 
as the alphabet itself was preserved in its integrity ; but when 
such a ligature as a had been degraded into ^, and then by the 
carelessness and haste of scribes had been levelled under e 
together with oe, and Latin c and ^ had come to represent two 
different sounds each — all this happening in Old French 
orthography — the phonetic tradition was broken, and spell- 
ing could only be half phonetic. 

281. The influence of Latin spelling in the Bomance 
languages — due, of course, to the continuity of the languages 
themselves — is shown not only in the retention of * soft ' c and 
^, but also in the later French 'etymological' spellings by 
which delte was made into debie with a ' silent ' 6, after Latin 
debitum. It is, however, doubtful whether this was done with 
any etymological intention — at least at first. Scribes who 
were continually copying texts written in an endless mixture 
of dialects would naturally seek refuge in the comparative 



70 HISTOEY OF ENGLISH SOUNDS. 

uniformity of the Latin spelling they were taught to reverence, 
and so would half unconsciously modify their unsettled French 
in the direction of the fixed Latin spelling. No doubt the 
pedants of the Benascence did attempt to ' reform ' spelling on 
etymological grounds, and occasionally with success, but nearly 
all the modifications of spelling that have been made in Europe 
since the introduction of printing have been phonetic, such as 
the dropping of silent e, the distinction between oa and oo in 
E. The 'reason why comparatively so few of the ceaseless 
attempts at similar reforms have succeeded, is that the early 
spelling reformers had not enough scientific knowledge and 
experience to grapple with the great changes in pronunciation 
and the corruption of the Roman alphabet. 

NORMALIZING. 

282. When we contrast the regularity of modem spelling 
with the irregularity of that of the Middle Ages, in which the 
same word may be spelt in half-a-dozen different ways on the 
same page, we are apt to assume that the older usage reflects 
the freedom of nature, the modern regularity being purely 
artificiaL But we soon find that such varieties as ME cutnej 
iume, come all mean exactly the same thing, and that where 
there are real underlying distinctions of sound, they are due 
to mixture of dialect — a mixture which, however, is often 
only apparent : the result of a scribe copying a ms written in 
another dialect which he only partially transliterates into his 
own. Another source of confusion is copying an older ms 
in an archaic spelling, which spelling, as a general rule, is 
neither retained nor discarded consistently, the result being 
more or less of an anachronism. 

288. The remedy for this confusion is normalizing, which 
takes one definite dialect, and selects one definite spelling for 
each sound, the result being a more or less absolutely uniform 
orthography, of which the ME Ormulum is one remarkable 
example, classical Sanskrit another. Normalizing has nothing 
to do with fixity of orthography. As we see, Sanskrit ortho- 
graphy was stereotyped together with the language itself, 



SOUND EBPBBSBNTATION. 7 1 

while Orm's spellings perished with their author. The present 
E. spelling, again, though fixed, is not perfectly normalized. 
Thus we denote the (ou) from OE a by -f « in stone^ hut by 
oa in moan^ although these two words have always had the 
same vowel from the beginning, and so on. 

SYNTHESIS. 

284. A normalized spelling on a rigorously phonetic basis 
will, of course, ignore such non-phonetic considerations as 
word-division, and will reproduce all the modifications which 
words undergo in different surroundings, as in the Sanskrit 
sandhi. It ought also to preserve the distinction between 
such doublets as (iSset) and (ISet). But in practice this is 
seldom done, it being found more convenient to write the 
emphatic form everywhere. The scribe, too, in writing has to 
pronounce each word to himself detached, and therefore in its 
emphatic form and free from such influences as sandhi 
and consonant-mutation. Of course, where variations in the 
form of a word are associated with marked divergencies of 
meaning, as in the Celtic mutations and such pairs as E. oney 
«(»), offy oft they are recognized in writing. 

286. This leads also to a general disregard of synthesis. 
Sanskrit denotes vowel-quantity everywhere, Greek only in 
some of the vowels which have distinct signs for the longs. 
In Latin the quantity is marked only by a diacritic which is 
generally omitted. Intonation is marked in Yedic Sanskrit 
and in some of the pre-classical Brahmanas. It was not 
marked in Greek till the Alexandrian philologists devised 
a scheme of accentuation for the benefit of foreigners. Hi 
modem languages quantity is often marked by doubling, as in 
Dutch, and less regularly in German and E., and stress by an 
acute accent, as in Spanish; this acute being primarily a 
mark of high or rising intonation, which was however — in 
Greek at least ^combined with stress. Our punctuation- 
marks seem to have been originally modulative, and a comma 
is still more or less equivalent to {^), though pxmctuation is 
now mainly logical. 



72 HISTOET OP ENGLISH SOUNDS. 

286. Word-division is disregarded in Sanskrit, though not 
in most Eastern languages. It was generally disregarded in 
Greek«and Latin, the division between words being marked — 
whenever it was marked — ^not by spaces, but by a point. In 
the early Middle Ages subordinate words — especially pre- 
positions — were generally run on to the following noun etc 
to which they belonged. The grouping of subordinate words 
round their centre was carried to a great extent in Old Irish, 
where, for instance, indfhirsin was written for ind fhir sin 
* of-the man this,' ' of this man.' 



INTERPRETATION OP SYMBOLS. 

287. The one essential difference between the phonetic study 
of living and of dead languages, is that the former are 
accessible to direct observation. But it is easy to exaggerate 
the importance of this difference. Even in studying living 
languages we are forced to rely mainly on the observations .of 
others, for no one can master more than a limited number of 
languages, and it is only the observations of a native that can 
be perfectly relied on, so that the statement of an old Sanskrit 
phonetician that, for instance, his to was formed by the lips 
and teeth is really worth more than an unphonetic German's 
analysis of E. to into u + ia^or an Englishman's statement that 
South German w is between w and v. 

288. The first means of determining the sounds of dead 
languages is, therefore, the direct statements of phoneticians, 
grammarians and others about them, whether in the form of 
simple description or of correction of assumed errors or 
vulgarisms. The results thus obtained may be supplemented 
by comparison with the sounds of other languages, and by 
phonetic transcriptions. 

238. Then we have the indirect evidence of the spelling, 
which is often as reliable as — if not more so than — the 
former. Such forms as i, i are, indeed, self-interpreting, and 
many others, such as lA^ nj\ though ambiguous in themselves, 
are often interpreted with certainty when taken in connection 
with other evidence, and with the history of the language and 



SOUND EEPRESENTATION. 7 3 

the general laws of sound-change. The very fluctuations are 
often instructive. Indeed, when we find the elements of 
a digraph liable to constant variation and transposition, 
we may be sure that this digraph is intended to represent 
one simple sound lying between its two elements, especially 
if it alternates with a single letter. Thus when we find the 
same sound written eo, oe, in ME, we may assume that 
it is meant to indicate some variety of {. The loss of a 
letter is, of course, often conclusive, as when OE Al becomes 
simple / in ME. So also are confusions, as when Late Mercian 
confuses y and i. 

240. The introduction of a new system of spelling often 
throws fresh light on a language, for each orthography 
brings out phonetic features of its own. Thus the distinc- 
tion of back and front c in OE becomes quite clear in the 
Frenchified spelling of the 13th century, in which the latter 
is written ci, 

. 241. The third great criterion is afibrded by metre. The 
evidence varies, of course, according to the nature of the metre. 
Latin verse enables us to determine with certainty the vowel- 
quantity, and OE and ME metre does the same to some 
extent. MnE metre enables us to determine the word-stress 
and to eliminate silent ^'s with considerable accuracy. The 
ornaments of verse — vowel-assonance, rhyme, and alliteration — 
also throw their own Ught on pronunciation. Here, however, 
we must be on our guard against those traditional influences 
which result in ' printer's rhymes.' Rhymes in the infancy 
.of the art are generally more or less imperfect, and even 
Italian never got so far as to separate close and open e and 
in rhyme, as Middle High German did. These imperfect 
rhymes — which may be printer's rhymes at the same time — 
such as lave : prove, are really * consonantal assonances.' 
Bhyme is especially valuable in reconstructing the dialect of 
the author of a poem, when it has been hopelessly disguised, 
as is often the case, by being copied from one dialect into 
another. Thus a Scotch poem, even if transliterated com- 
pletely into Southern English, would still betray its origin by 
such a rhyme as home : na»f^= Scotch hame : name. 



74 HISTOET OP ENGLISH SOUNDS. 



ARIAN SOUNDS. 

242. The following is a classification of the chief Arian 
languages according to their relationship. 

(a) East- Arian or Asiatic : 

(i) Sanskrit. (2) Iranian (Zend and Old Persian). 

(b) West-Axian or European : 

(3) Greek. 

(4) Latin. (5) Celtic. 

(6) Slavonic (Old Bulgarian). 

(7) Baltic (Lithuanian and Lettish). (8) Germanic. 

248. It will be observed that Gk and Lt have nothing in 
common except that they are both West- Arian, that Celtic is 
most closely allied to Lt, and Gmc to Baltic, Gmc lying 
geographically between Baltic and Celtic. 

244. The development of these languages seems to have 
been the result rather of a gradual divergence than of an 
abrupt separation, although no doubt the latter process may 
often have hastened the divergence. Indeed, if all the 
Ar. languages had been preserved, we should probably find it 
difficult to draw any definite line between the difierent 
groups. As it is, Armenian seems to be really a link between 
Iranian and Slavonic, and therefore between Asiatic and 
European, and Albanian may turn out to be a similar link 
between Gk and Lt. 

246. By comparing the separate languages in their oldest 
forms, and collecting those resemblances which could not have 
developed independently, and must therefore be due to 
community of origin, we are able to reconstruct parent Ar. 
with some certainty — at least in its main features. It was 
a highly inflectional language, complex and yet symmetrical 
in structure, with a rich sound-system, which, as regards 
the vowels, is very faithfully reflected in the oldest Gk, 
the general structure of the language being otherwise best 
represented by the oldest Vedic Sk. It bore a striking 



AEIAN, 75 

resemblance to MnE in its extreme sensibility to stress- 
influence. 

246. But parent Arian shows distinct traces of an earlier 
pre-inflectional stage, in which sentences were made up of 
indeclinable words or ' roots,' whose relations to one another 
were expressed partly by position, partly by the addition of 
shortened words which by degrees became incorporated into 
the preceding root- word, inflection' being the result. The de- 
velopment of inflection implies complete subordination of one 
word to another ; but it is possible for two words to be indis- 
solubly joined together, each retaining its full individuality, 
as in hand-made. Such compounds as Adsta-trta * hand-made ' in 
Sk, kheiro-poUtos in Gk are, in fact, nothing else but fragments 
of pre-inflectional sentences, as is shown still more clearly in 
the Sk copulative compounds, such as ahd-rdtrdm ' a day and 
night,' which in some cases even take an independent accent 
on each member. Inflection in all languages is developed 
mainly in connection with other words in a sentence, and 
words forming sentences by themselves never developed 
inflection at all; hence we have pre-inflectional words in 
vocatives and imperatives, such as Sk deva 'god I' bkdva ' bel' 
It will be observed that some at least of these roots were 
dissyllabic. It is probable that the Ar. monosyllabic roots 
which we see in Sk m£=Gk dps contrasted with dgvas =Gk 
hippos^ are really unemphatic forms, which originally existed 
side by side with the fuller emphatic ones. 

247. The development of the Ar. vowel-system cannot be 
understood without a knowledge of Ar. accentuation. That 
the free accent of Yedic Sk should be, in the main, that of 
parent Ar. is in itself very probable, and is made certain by 
Yemer's law (315), which explains certain irregularities in the 
Germanic consonant-shift by the position of the accent in Sk, 
showing that parent Gmc and Vedic Sk must have had 
a practically identical system of accentuation which can only 
be the result of common origin. 

248. There are three accents in Sk : uddtta (raised) =' acute,' 
anuddtta (unraised)=' grave,' and n;arti^s' circumflex.' The 
acute is the emphatic accent, and was either a rising or 



76 HISTORY OF ENGLISH SOUNDS. 

a high level tone. The syllable immediately following an 
acute is always circumflex — that is, probably a falling glide- 
tone — unless it is itself followed by an acute, in which case it 
is grave : iena but tena te. Every syllable before an acute or 
after a circumflex is grave : ahdm^ hhdvami. The acute was 
no doubt accompanied by stress, for the frequent dropping 
of grave vowels can only be explained as the result of want 
of stress. 

240. In parent Ar. every vowel had a different form under 
the different accents. The result was a variety of vowel- 
series, each with the three stages,«/re^»^,m^t2t»»t,and weak. The 
most important of these is the e-o series, which is evidently 
a modification of original a. Under the acute accent a became 
e (through x), under the circumflex it became o (through j), and 
under the grave it was dropped altogether. The first two 
changes, which are evidently acoustic (133), are shown in Gk 
Mppoa (where i is a later modification of €\ Lt eqvtis (older 
eqvos) = Sk dgvaa from Ar. ^ecwos 'xQOjS^ (pre-Arian *dcicai). 
It will be seen that Sk opposes a uniform a to the Gk, Lt 
(and general European) e, 0. But that this is only a com- 
paratively late levelling in Sk itself is proved by a variety of 
facts. Thus Ar. k q» become ^ q in Sk before Sk a=Ar. and 
European ^, as in cfl=Lt qne (Ar. ^ke)^ but not before Sk 
«= Ar. 0. European in open syllables, as in Gk gSnu *knee,* 
ph6rod * tribute,' is represented in Sk by a, as in jdnu m3*'5i, 
bhara do3«(d3 ; European followed by two conss. being repre- 
sented in Sk by short a, as in daddrga aj3®3^^3 * ^ saw* = Gk 
dedorka. This seems to show that the circumflex had the power 
of lengthening a vowel under certain conditions (when 
followed in parent Ar. by a vowel with a grave accent ?), the 
short European in gdnu being due to some analogical 
influence. But we find also an European in the ^-series, as 
in Gk jihor * thief,' connected with phero^ klqps * thief,' con- 
nected with kl^j)td * steal.' 

260. Under the grave accent the vowel is dropped entirely. 
Thus Sk kdrdmi 'make' (a=Ar. e) has the past passive 
participle krtd. So also kdlpdmi ' arrange ' has partic. ilptd. 
In the other languages these syllabic liquids have been 



ARIAN : ACCENT. 77 

resolved into non-syllabic r and / accompanied by a distinct 
vowel. Thus Qk dhkomai 'see' has aorist Sdrahon—^s. 
ddrgam. The original syllabic nasals have not been preserved 
even in Sk. Thus Ar. ^IntS 'stretched* appears in Sk as 
taldy in Ok as tatds, in Lt as tentas, 

261. Words with the diphthong ei in their strong stage, such 
as Ok eimi ' I go/ Sk emi, show simple i in weak forms such 
as the plur. imen ' we go ' (with shifting of the Arian accent) 
= Sk im(U. It is evident that the treatment of the diphthongic 
vowel is perfectly parallel to that of the liquids, ei being 
equivalent to e;\ which is parallel to er and el. So also the 
strong eu is weakened to i^ by dropping the «, as in Gkpustds 
' known/ pres. peitliomai. 

262. The reduction of er to r, of m' to « was, of course, 
a gradual process, and there must have been many inter- 
mediate stages. When we find Omc eunu ' son ' contrasting 
with Sk sunu^ Ok Uo9 ' life ' with Sk jivd ' alive,' it is natural 
to suppose that the long vowels really represent an older stage 
of weakening than the short ones. It is probable that sunu 
and 9unu etc existed side by side in parent Ar., the latter 
being, perhaps, the more emphatic form. This suggests a 
similar coexistence of r and f (long syllabic r), and when we 
find Sk piirnd of wic] « full ' (literally * filled ') with ur instead 
of the r of krtd^ and Ok atrotSs ' strewn ' with ro instead of ra^ 
we canfiot help inferring Ar. *jjfnS, *em6. 

268. When e is fianked by unvowellike consonants, especi- 
ally stops, it is generally kept in the weak stage; thus 
in Ok the weak eheptSa 'seen' has the same vowel as the 
present dcSptomau But it is also dropt, as in the Ok aorist 
infin. ptSsihai^ pres. pitomai * fly,' Sk Atmi (Ar. */*i»i), plur. 
wndn. The e was probably dropt everywhere at first, and 
then restored by the analogy of the strong forms. Perhaps, 
however, such weak forms as *ihpU and ahepU may have 
existed side by side parallel to 9unu and aurvi etc. 

264. In Sk many words ending in a cons, show accent-shift 
in inflection, thus v&k ' voice' has ace. vaeam^ gen. vdca%y emi ' I 
go ' has plur. imds ' we go.' So also in Ok dps =:Sk. vdk has ace. 
6pa^ gen. opda. There is no shifting of accent in the inflection of 



78 HISTORY OF ENGLISH SOUNDS. 

such words as dgvaSy Mppos, but there is every reaaon to believe 
that the later uniformity is not original. When we find strong 
eu in an unaccented syllable in Gk leukSi^ gen. leuioH ' white^' 
and, conversely, weak r accented in Sk vfia * wolf,' we see 
that in the Gk word the original change *l€ukoSy gen. ^lukSso 
has been levelled by a compromise between the vowel of the 
nom. and the accent of the gen. In Sk, on the contrary, the 
accent of the nom. *verkos has been associated with the vowel 
of the gen. ^vrkho. Gmc ^wulfa (OE toulf) points also to Ar. 
^tDfko. The Zend vehrka^ again, preserves the vowel of the 
nom. So also OE 9wrfn ' sleep ' and Gk Mjmos point to Ar. 
*9wSpnos^ ^supn^so, toe being weakened into u in the same way 
as eu (z=ew) mpeuthomai. 

266. The following are the main types of the ^-series in 
their three stages : 



rang 


medium 


weak 

• 


ek 


ok, ok 


ek, k 


er 


or, or 


f,r 


en 


on, on 


n, n 


ei 


oi, oi 


i|i 


eu 


ou,5u 


u,u 



266. The other series are less clear. The a-series has a in 
the strong stage, as in Gk dgo * drive,' Lt ago, Olcel. aka, Sk 
djdmi=^AT. *dp, Gk ailAo *set fire to,' OE dd (froin Gmc 
*aida) ' fire,' Gk Aaud *dry.' The weak stage is quite parallel 
to that of the ^-series : Sk jmdn * path,' Gk pass. ptc. epaktSsy 
the former representing the ' short-weak ' (as in pUsthai) the 
latter the * long-weak' stage (as in Sk skeptds); iddkd < burnt,' 
*pure,' Gk UAardsy OE idel *idle' (originally *pure,' * empty'), 
the latter being parallel to Sk jivd. 

267. The <?-series is represented only by a few words, such 
as Gk dzei 'smells,' Lt olOy Gk ano^d 'open,' krado 'strike.' 
These examples represent the strong stage. The long-weak 
stage is shown in Gk optdon 'to be seen,' aniikru 'against' 
(literally 'striking against'), the short-weak stage in the 
variant antikm, 

268. Some words have the long vowels a, e and o in their 



ASIAN: VOWELS. 79 

strong stages, as in the Ok yerhs pAdmi 'speak/ hUtdmi 'stand,' 
iithemi 'place,' diddmi 'give.' In all these series the short- 
weak stage drops the vowel altogether, as in Sk devdtia=. 
*deva-dta * god-given,' pres. Sk ddddmi=Qli d4ddmu The 
long-weak stage has i in Sk, a in European, as in the prt. 
pass. ptcc. Sk atkitdy Ok HatSa from Mstdmi^ Gk phamSn * we 
speak,' Lt datus 'given.' This European a=Sk i may point 
to an Ar. p X* 

260. In many cases the long vowels appear to be lengthen- 
ings of short vowels in the e, a, OHseries. Thus the e of Lt 
pes * foot,' the o of (Doric) Qk pos, Gothic Join appear to 
be lengthenings of strong e and medium o respectively (cp 
Lt pedes, Gk pSdes). So also the a and o of Gk stratdgds 
'army-leader,' Sdode 'smelt' belongs to the a and o-series 
respectively. 

260. Where o appears in the a. and ^-series, as in Gmc 
^siola (OE stol) ' stool ' ( + Gk Mstdmi), it may represent the 
medium stage of these series, being the result of circumflexing. 
Indeed, Gk bomds ' altar ' stands in the same relation to bdtna 
' step ' as tormds ' log ' does to k^rTJia ' anything cut small.' So 
also Gk rhegnumi ' break ' perfect (rrhoga is quite parallel to 
leipo ' leave ' perf. Uhnpa. 

261. Whatever its origin, the Ar. vowel-system must have 
had somewhat the following form : 



' i e o I e 



1 


u 


a 


e 







ei 


oi 


ai 



a 



ai ei 01 ai ei oi 

au eu ou au eu ou 

262. These vowels are represented as follows in Sanskrit, 
Gk, Lt, and Gmc : 



8o 



HISTORY OP ENGLISH SOUNDS. 



Ar. 


a 




e 


u 





a 


- 


e 


u 


6 


Sk 


a 




a 


u 


a 


a 




a 


u 


a 


Gh 


a 




e 


XI 





a 




e 


u 


6 


Lt 


a 




e 


u 





a 




e 
e 


u 


o 


Gmc 


a 




e 


a 


a 


6 




u 


d 



In OBg a is represented by o^ i by < I, « by ii( 1, « by y I, 
ohy a. In Lith. o is represented by a, and a by o — both as in 
Gmc — and ^ by 6 (=oaa). 



Ar. 


ai 


ei 


oi 


an 


eu 


ou 


Si 


e, ay 


e, ay 


e, ay 


6, av 


6, av 


6, av 


Gh 


ai 


ei 


oi 


aa 


eu 


ou 


Lt 


ae 


1 


i 


au 


u 


u 


Gmc 


ai 


i 


ai 


au 


eu 


au 



Sk e^ appear as ay ](d, av 3^ before vowels. 

268. The correspondence of the long vowel diphthongs is 
not certain. In Sk the first element of all of them necessarily 
becomes a. In Gmc it seems to become a, so that di^ ei, ai are 
all levelled under ai. 

264. In OBg all the diphthongs are smoothed, au becom- 
ing u. 

We will now consider the vowels more in detail, giving 
examples from the different languages. 

266. a. Gk dffo, Lt a^ere, Olcel. aia 'drive.' From the same 
root Sk djras, Gk affrds, Lt affer, Goth, air, OE acer * field.' 
Gk arSoy Lt ardre, OBg oratiy Goth, atyan^ OE ^rian * plough.' 
Sk dpa, Gk ajxf, Lt ab, Goth, of, OE tf. 

266. i (weakening of ei). Sk bibhidimd^ Lt fidintM^ OE biton 



arian: vowels. 8i 

* we bit.' So also in most of the other verbs in Gmc with 
i in the present. Sk vidmd^ Gk idinefi, OE witon ' we know.' 
Sk iddm, Lt id, Goth, iia, OE Ait ' it.' 

267. e. Sk bkdrdmi, Gk phSrein^ Lt /err^, OE i^aft * bear/ 
Sk mddhuy Gk mSthu, OBg m^Ji^, OE «»&^2» 'mead.' Ar. c; 
was probably very open { = x?), as it returns to a in Sk. 

268. u (weakening of eu), Sk hihudhimd, OE hudon 'we 
announced.' Sk buddhd^ Gk pustSa, OE doi/^ ' made known.' 
So also in the other Gmc verbs with eu in the pres. 

268. o. Sk asfaUy Gk oklOy Lt octo, Goth, ahtau ' eight.' Lt 
nox, Gk ^1^^, Goth, nahis ' night.' o, the medium stage of e, is 
seen in perfects such as Gk dSdorha, Sk daddrga, Gk pres. 
derkamai. So also Gmc a in bar 'bore'=o, Sk babhdra with 
lengthening (249). Goth. ^a^'a» 'set' = Sk sdddydmi, from 
^<2-. Gk khSrtos, Lt Aortus, Goth, ^ar^b ' yard ' ; the e is seen 
in Gk eukheres ' easy to handle.' 

270. &. Sk Hxfidmi, Doric Gk Matdmi (Attic Aislemi)^ Lt 
stare, OBg ^^^i, Lith. ^^^i ' stand/ OE «/^^ ' stool.' Gk mater 
Lt mater, OE ^5rf^ * mother.* 

271. i. Sk Jivd, Lt r^vw* * alive/ OE cifi * sprout.' Short i ill 
Gk bios, OE wi<? (266). 

272. e. Sk dddAdmi, Gk (Doric and Attic) tUAemi 'put.' 
OBg dtti * do,' Goth, rf^rf* ' deed.' Gk Aiemi * throw/ Lt semen, 
Goth. «^^ * seed.' As Ar. ^ returns to ^ in Sk and in some 
Gmc languages, it probably had the open sound jf. 

273. u. Zend srHto * celebrated ' (cp Sk grutd * heard,' Gk 
klutSs with short vowel), OE Alud * loud.' 

274. o. OE do 'do/ dom 'doom,' Gk tAomds ^heap,' connected 
with tUAemi ^put/ Gk molos 'trouble,' OE mctpe 'weary' 
(from mdfi'), 

276. ai. Gk aitAo ' bum/ OE dd (from *aid) ' fire.' Gk laids 
' left,' Lt laevus, OE slaw (from ^slaiw) ' weak,' ' slow.' 

276. ei Gk leipo 'leave/ Goth. leiAvan (ei=i) 'lend.' Gk 
steikAo, OE stlgan 'ascend/ Sk emi, Gk eimi 'I go.' Gk 
dedmUmi 'show/ Lt dlco 'say/ OHG 2^z^i;» 'accuse,' MnO 
verzeiAen ' pardon.' 

277. oi. Gk oidos * swelling,' Lt aemidus ' swelling ' adj, 
Gmc ^aitra (OHG eitar, OE dtor 'poison'), oi, the medium 

a 



82 



HISTORY OF ENGLISH SOUNDS. 



stage of eiy appears in Sk eman^ Gk oimos ^ path' (op ei in Gk 
eimi ' I go '). Sk rireca, Gk l^loipa ' I left/ Goth. laiAv? ' lent' ; 
Gk loip6» ^ remaining/ Goth, laiba ' relic' 

278. an. Gk (Aeolic) auM^ Lt aurora *dawn/ OIcol. awtr 

* east.' Gk paud ' cease/ lApauciu, Goi3i.favai * few.' 

278. eu. Sk bodhami ^ watch/ Gk peitkomai < enquire/ Gmo 
leudan (Goth, biudan) ' offer.' Gk geud [i=i*geusd) ' taste/ Goth. 
kiusan 'choose.' 

280. ou. Sk perf. bubAoda, Gmc *baude (Goth, bauf) 'offered' 
Gk perf. eileloutka ' came.' 

281. Of the diphthongs with the fbrst element long a few 
examples must suffice. Arian ei is seen in Sk prdyas, Gkj)le(dn 
(^=z*j)le{dn\ Olceh Jleirt 'more.' Gk pleistos 'most' («=^i), 
Olcel. flestr (^=^'= Germanic ai), Arian &u is seen in Sk 
dj/au9 ' sky/ with which cp Gk Zeu^, The Ionic Gk eo% (Attic 
Mds, Eolic aud9) points to Ar. ^atLSos. Sk gaut plur. gdvas 

* cow/ Lt io*, Gmc * kd points to Arian ou. 

282. The syllabic liquids are represented as follows in the 
four principal languages : 



Ar. 


r; 1 


f 


n 


h 


Sk 


r;l 


ir, UT 


*an, a 


a 


Gk 


ar, al ; ra, la 


ro, lo 


'an, a 


a 


Lt 


or ; ol, ul 


ra, la 


en 


an 


Gmc 


or; ol 


ar,al,ra 


'in, un 


an 



The forms marked (•) develop only in syllables which in 
later Ar. came to be accented. Syllabic m can hardly be dis- 
tinguished from n, 

283. p. Sk vria, OHG wolf. Sk sMd ' spread ' ptc, Gk 
stratSs 'army/ Sk vrddhd 'grown/ Gk blastSi 'sprout.' Sk 
hrd- ' heart,' Lt cord-. 

284. p. Sk jlrndm ' ground ' ptc, Lt grdnum ' grain.' Sk 
irmdy Gmc arm. After lip conss. f becomes f^r in Sk as in 
purvidy Gkj>roioa ' foremost/ QoHi./rauJa ' lord.' 



ABIAN : CSONSONANTS. 



83 



285. The different languages vary in the length of the r. 
Thus to Sk sMd corresponds Gk strolds, Lt strdtus, both 
pointing to Ar. f . To Lt grdnum corresponds Gmo kam {or 
from r), to SiLpurnd * filled ' OHO vol. 

286. n. Sk sdnti^ OE rind (from * sin}) 'they are.' Sk 
taidy Gk taids, Lt ienius, 'stretched' (cp the Gk pres. tefnd 
from ^ tdnjd). Sk gatd, OE hund ' hundred.' 

287. n. Sk gdtd ' gone ' from *g7htd or *ghtd^ Gk bdtH * go ' I 
Sk dii * duck/ Lt aftas^ OIceL ^ (from * andu). 



CONSONANTS. 

288. The following was the Ar. consonant-system : 





bach 


front 


forew. 


lip 


open 




J 


r; 1. s,z 


w 


nasal 


n 


n 


n 


m 


stop 


k, g 


C.J 


t,d 


p,b 


aspir. 


kh,gh 


ch,jh 


th,dh 


ph, bh 





d> 


(D;00.S,8 


9 


J 


L 


7 ' 


r 


Q9, GP 


affi 


0, CD 


D, D 


Qo^, Q00 


Qo, me 


DO, <W 


D*>, DO 



The back and front nasals occur only before back and front 
stops (and aspirates) resp. in such combinations as ng^ nc. 
z occurs only before voiced stops in the combinations zg^ zdh 
etc. 

289. The breath aspirates th etc were no doubt stops fol- 
lowed by a stressed breath-glide, as they still are in Lidia. 
The voice aspirates dh etc are described by the Sanskrit 
phoneticians as voiced stops followed by sonant breath, which 

o % 



84 



HISTOET OF ENGLISH SOUNDS. 



may mean either a mere emphasizing of the following voice- 
glide or a distinct guttural croak. In Yedic Sk d mc between 
vowels becomes the inverted / ooc, and consequently ^i becomes 
fA, which is expressed not by a single letter, as is the case with 
the other aspirates, but by /+' sonant A,' which in this case 
could not well have been anything but e. Both classes of 
aspirates were originally nothing but emphasized stops, whose 
off-glides were exaggerated. 

290. There appear to have been two ^-sounds, one (d) ?) re- 
presented by z in Gk, as in zn^dn, Sk yu^dm ' yoke,' the other 
(x ?) being weakened to a mere breath in Gk, as in the pro- 
nominal Ads=zSk yds. In Sk the latter ^ in reduplication- 
syllables becomes i- as in iydja perf. of yaj ' sacrifice,' instead 
oiya-. So also in Sk some verbs reduplicate with ra-, such as 
vardA ' grow,' some with i^-, such as vae ^ speak,' pointing to an 
analogous distinction of s and %. 

291. The distinction between r and / certainly existed in 
Ar. (although in Zend both are represented by r, and they are 
not separated so strictly in Sk as in European), but probably 
in a different form : it is possible that r was represented by 
trilled, / by untriUed r. 

292. The following table will show the development of the 
back and front stops in the different languages : 



Ar. 


k 


g- 


gt 


c 


J 


Jb 


Si 


k, 


g>j 


gh,h 


? 


• 


h 


Zend 


k, c 


gj 


• 

g>J 


s 


z 


z 


Gk 


k, p, (t) 


g.b,(d) 


kh, (th) 


k 


g 


kh 

Kg 


U 


q, c 


g 


b,g 


c 


g 


OBg 


k, 2, c 


V 

g. z 


g.z 


s 


z 


z 


Lith. 


k 


g.^ 


g.2 


8Z 


2 


i 


Gme 


hw,h(w,g) 


kw 


gw, (w) 


b(g) 


k 


g 



GERMANIC. 85 

Sk (?=Q, y=(D, f =0 (or 8^?), ^=0. Qk ^^=00. Lt ^=0^ 
(?=a OBg <!f=02, i=e. Lith. i^=e, *i?=e, 

298. The At. breath aspirates M, tiy pi are distinctively 
preserved only in Sk, having run together with original ^i 
etc in Gk. In Gmc they were confounded with Ar. i, t, jo, 
which themselves were aspirated into M etc, passing after- 
wards into open conss. (3 13). 



GERMANIC SOUNDS. 

204. The Old Qermanic languages fall into two main 
divisions : 

(a) East-Qermanic : 
(i) Gothic 

(2) Scandinavian (Icelandic, Danish, Swedish), 
(i) West-Gtermanic : 

(3) Low-German (Old-Saxon, Old-English, Frisian). 

(4) High German. 

Within Low German E. and Frisian again form a special 
group ' Anglo-Frisian.' 

206. By a comparison of these languages among them- 
selves and with the other Ar. languages we can reconstruct 
parent Gmc with some certainty. This pre-historic Gmc 
language differed from its extant descendants in two im- 
portant features. It stiU kept the free Arian accent, often 
shifting from one syllable to another in different inflections 
of the same word. Thus in the verbs the pros, and infin. had 
the accent on the root syllable (b^an, birijf), while it was 
thrown on to the end-syllable in the past partic. (borand)^ 
and the 2nd sg. and pi. pret. (budun). Nouns in -i and -» and 
weak masculines also throw it forward (fatmind'fszOE gemynd^ 
9un4, bogS^OE boga). Afterwards the accent was laid uni- 
formly on the first syllable, which was generally, though not 
always, the root syllable. 

206. It also had a complicated inflectional system, the verb 
having had a number of tenses which in GoUiic are reduced 



86 HI8T0ET OF ENGLISH SOUNDS. 

to two — present and preterite. One of the most marked charac- 
teristics of the Gmc languages is their striving after symmetry 
and regularity, of which their vowel-gradation in such forms 
as sififf, sanff, mng — a system which was built up out of the com- 
plicated At. vowel-system by a slow process of simplification 
and analogy — ^is an example. This may be partly due to 
the influence of the Ugrian languages, with which the Gmc 
came into dose contact for many centuries, just as the want 
of symmetry and isolating tendencies of Celtic seem to be 
due to the influence of a language of the Basque type. 



VOWELS. 

297. The following is the Gmc vowel-system : 

{i u . f i u 

a-^ _ . 
e o ( 63, e o 

. r au 
ai \ 
\ eu 

208. a=Ar. a (akr) and o (nakt, gard). 

299. i=Ar. i {toitan) and e. Ar. e became i in Gmc before 
nasal + cons., as in OE Undan compared with Aelpan^ and 
before an t or y in the next syllable, as in OE hiripy bir{e)Jf 
' bears ' = Sk bkdrati, Ar. bh^reti. It is possible that these two 
t's differed in sound (I and I ?). Unaccented e seems to have 
become i everywhere in later Gmc. 

800. u=Ar. u (OE budim^ sunu). Earlier Gmc o becomes u 
under the same conditions as e becomes t, as in OE bunden 
compared with holpeUy gylden (from ^gulpina) from gold. 

dOL e=Ar. e (beran). In some wordB it is Ar. i mutated 
by a following back vowel, as in OE nest from *nizdd. 

802. o=:Ar. u mutated by a following back vowel, as in 
.OE coren * chosen,' Gmc ^kozand (cp Gk geuo). This influence, 
which is only occasional with Ar. e, is regular with u. Cp 
the OE partico. stigen^ togen etc. A following u^ however, pre- 
serves original t;, as in OE bvdun, budon 'they offered' etc* 
Another main source of Gmc o is the development of a para- 
sitic vowel out of the Ar. syllabic liquids (OE com), which 



GERMANIC: CONSONANTS. 87 

before n becomes u (OE Aund). Such particc. as OE boren 
point to *brrand, equivalent to *bfand, witii the f resolved into 
syllabic r+ consonantal r. 

303. ft. There was no pure a in Omc, as Ar. a became 
0, but in the combination ^anA jjc from Ar. ani the n nasal- 
ized the preceding vowel, lengthened it, and was then 
dropped, as in the Ooth. preterite JidAta, OE j^Ale, ^Ate, 
whose points to earlier nasality. 

304. i= Ar. i (OE cip) and ei {stlgan). Nasalized i from *inA 
(=Ar. ink^ eni), as in OHO liAti 'easy,' 'light' (Sk lan^A 
• leap '). 

305. 11= Ar. u (OE Alud). Nasalized u from *itnA (= Ar. nniy 
older Gmo oni), as in OE J^AU ' seemed,' pres. fyncan from 

306. »= Ar. e (OE d€ed\ which probably had the same broad 
sound -[«. In Scandinavian, OSaxon and High German a 
became a (OIceL ddj)^ OSaxon dad, OHO ia4). 

807. e. A vowel of obscure origin in such words as Aer 
' here ' ; sometimes the result of contraction. 

308. o=Ar. a (OE moder) and (dam). from Ar. a prob- 
ably had at first a broader sound than original 0. 

309. ai=Ar. ai {aid) and ai {aitr, laiba). 
810. au= Ar. au (austr) and ou (baud). 
311. eu=Ar. eu (beudan). 



CONSONANTS. 

312. The following was the Gmc consonant-system : 

h,x j r;L J>,«; s, z w. f , v 

n n m 

k, g t, d p, b 

A=rC, J = €,/=0, f;=3. 

313. The most prominent feature of the Gmo as compared 
with thjB Ar. consonant-system is the Gmc consonant-shift 
(Grimm's Law, lautverschiebung), by which the Ar. breath 
stops (and breath aspirates) become open conss., while the 
voice stops are unvoiced, and the voiced aspirates become 



88 HISTORY OP ENGLISH SOUNDS. 

simple voiced stops; the following being, for example, the 
correspondence of the point series : 

Jr. t, th d dh 

Gmc J> t d 
Sk tudm^ Gk td, Lt tu=tQoih. Jfu *thou/ Sk. pAena=^0'E f dm 
'foam/ Sk ddmi, Gk ^do, Lt edd^OE etan 'eat.' Sk mddhuy 
Gk metAu=^OiE medu 'mead/ 

314. Of these changes that of d into t was evidently more or 
less direct (through whisper). That oitioj) cannot have been 
by direct opening (as in Olrish athir^Olk pater), for in that 
case Ar. d would have become 9; it must, therefore, have been 
through the aspirate o^, Ar. t and tA thus running together. 
dA in the cognates shows two lines of development. In Gk 
the glide is unvoiced and the resulting Q}o naturally becomes 
GO. In Zend, Slavonic, Baltic and Celtic the glide is simply 
dropped, dA becoming d. It seems natural to assume that 
Gmc followed the same course as its neighbours. As regards 
the order of the changes, it is clear that dA could not have 
become d till Ar. d had become t, and that this latter change 
could not have taken place till Ar. t itself had been modified — 
otherwise some two of the three must have run together. 
The changes must, therefore, have begun with that of t into/ 
through tAy d then taking the place of Ar. t, and, lastly, dA 
taking that of Ar. d. 

This general scheme is, however, modified in detail by other 
laws. 

315. The most important of these is Vomer's Law, by which 
original Gmc A, /, / (from Ar. i, t, p) became ^, ^, v in syl- 
lables which in Ar. were unaccented, s becoming z under the 
same circumstances, g, 9, v afterwards were stopped into 
g, dy i, and z became r through (D\. The ff, d, h which arise in 
this way are called 'weak,' to distinguish them from the strong 
g etc from Ar. gA. Hence the so-called * grammatical change ' 
by which HeuAan (OE teon) ' pull,' *8eufan (OE seopan) ' boil,' 
^Musan (OE ceosan) 'choose' ^Aafjan 'raise' have their past 
particc. *to^and (OE togen), *so9and (OE soden), ^kozand (OE 
c(yr€n\ *Aavand (OE Aafen), In these examples the unaccented 
syllable containing the breath open cons, is separated by an 



GERMANIC; CONSONANTS. 89 

intervening syllable from the accented one. If the cons, is in 
contact with the accented vowel, the rule is that the breath 
open is preserved if the accented vowel precedes, as in ^tSukan^ 
^hrojiar (OE bropor) =Sk bhrdtdy voiced if the accented vowel 
follows, as in ^fwfSr (OE/^rfer)=Sk pitd, ^iogS (OE toga) 
' leader.' As this weakening is due to the voicing influence of 
the surrounding vowel, it does not extend to the beginning 
of words. 

816. The next exception is Paul and Kluge's Law, by which 
early Omc gg, dd, bb, became ki, tt, pj), as in OE ^rnocc (early 
Qmc *smuggd) compared with smugan ^ creep.' These double 
voiced stops arose from Ar. -Ikn, -ghn with the accent on the 
following vowel, the weak n being assimilated to the preced- 
ing consonant; if, on the other hand, the accent is on the 
preceding vowel, the n is preserved. The change of Ar. -In 
into Qmc U foUows the same law: op Olcel. gin 'ell' from 
Ar. *6lnd (Lt ulnuij with full from Ar. *pm6 through Gmc 
*/b6i4, ^foM ; so also OE tculle from Ar. wlno. In the case 
of the stops + n the order of the changes was as follows : 
Ar. &kna ^hna aknd aghn£ (ftgi^) 
Gme ^Lhna ^igna ahnA agni. akn& 

a^ni 

agn& 

^na igna akldl akkil akk6 
Examples are : Goth, anhns < oven ' from *4kno9 (Sk ukhd < pot '), 
OE 9wefn ' sleep ' from *9wSjmo. OE liccian (cp Goth, laigon 
* lick ') from ^lijAnd' (Gk likAneuo, OBg lizati) through *lignd, 
^liggd; OE smoce from ^smuhtd (OBg imyiati */ * creep') 
through ^smugnd^ ^smitgnd, ^smuggd; OE Aoppian 'hop' from 
^iupnd (OBg fyp&i) through nuvnd, nubnd, Hubbd. OE 
hce 'lock of hair' from *lugnd 'bent' (Gk lugdd 'bend') 
through *luind, ^lukkd. After long vowels and diphthongs 
these double conss. were shortened. 

317. Some consonant-combinations show special develop- 
ments. Ar. sk^ 9t, sp remain unchanged, as in OE JUc=Lt 
pisciSf steorra=^Gk aster ^ Lt stella, ipiican ^Lt spud. Ar. zd 
becomes ^, as in OE nest = Sk nldd (from *nizdd, from 



90 HISTORY OF ENGLISH SOUNDS. 

^ni-s{e)d6 * sitting-down'), Lt nidus. Ar. z^A, zdh become zg^ 
zdy as in OE mearg (Gmc ^mazga) T=iOBg mozgH^ Sk majjd 
'marrow/ Goth, mizdo (OE meord)^0!\L mistAds, Sk midAd 
* reward.' 

818. 'Ar. ity pt becomes At^ ft (through OPO^ D<>o), as in Goth. 
a^^f»=Gk oktOy Aaft * captive ' =Lt captus. Ar. U (UA) becomes 
regularly m (through o«>o, djo^ uv;), as in OE geunss * certain ' = 
Gk ui^ from Ar. ^wittS (from *widtd)y sometimes 9t by ana- 
logical and other influences, as in OE toast 'thou knowest,' 
Goth, vaist (Gmc *wais9)^ Gk oistAdy SkvettAa^ this toast etc 
owing its ^ to the analogy of Gmc tnaAt ' thou mightest ' etc 
This ss was shortened in Gmc after a long vowel or diphthong, 
as in *Aaisi (OE A^s) ' command ' from *Aaissi, *Aaitti (cp Goth. 
Aaitan ' command '). So also OE mos ' food ' from ^mossa, *matta 
(cp Goth, fnat 'food'). 

dl9. The variation between k and ^ etc in such forms as 
OE sucaHy sugan 'suck,' idcmg 'pirate,' idg 'war,' is often due 
to the influence of a lost nasal in Ar. In Ar. a nasal voiced 
a preceding breath stop under conditions which have not yet 
been determined (perhaps originally only when the syllable 
containing the stop was unaccented), as in Sk rgmfn 'singing' 
{re 'song'), Gk mignumi 'mix* (Sk migrd 'mixed'), Lt mendax 
' mendacious ' {mentiri vb), whence in Gmc a parallel alter- 
nation of A {£) with k etc. The probability of nasal influence 
in the case of voicing is confirmed by the cognate Lt vincere. 

820. The development of the Ar. back stops shows some 
peculiarities. The original representatives of Ar. k {iA\ g, gh 
in Gmc are AtOy iw, gta with the original Ar. rounding. But 
in an early stage of Gmc, in which Ar. o was still preserved 
as well as Ar. ^, u, u, the to was dropped before these round 
vowels, while it was kept before a, i, e. Thus Gmc ^Aawtoan 
(OE Aeatoan) ' hew ' corresponds to Ar. ^Aato- (Lt cudo, OBg iavq)^ 
^ko 'cow' (OE eu) to Ar. *gd (Sk gdtAs^ Lt bds\ but ^Avilo 
' time ' (OE Ami) to Lt qvietus^ OBg poilti ' rest,' Olcel. Avas% 
'sharp' (from ^Avattd) to Lt eattis^ Goth. Aven 'woman' to Ar. 
^geni (Sk Jdni), The to is not developed initially before a 
cons.: OE Alc^ 'bread' (Gmc Alaibd)=:IAih. ilepas. In weak 
Gmc syllables (31 j) A10 becomes gw^ which drops its to before 



GEBMANIO: CONSONANTS. 9 1 

original and «, as above, for example in Ooth. fairguni 
(ai=Gmc ^)=Lith. Perhinas (accent originally on the e). 
The to was, of course, kept before original a, f , ^, but here the 
g was dropt in accordance with Siever's Law, by which every 
gw — ^whether from Ar. k or gh — becomes w in weak Gmo 
syllables, as in Qoth. naus ^ corpse ' from ^4ia(^)«t^, Gk nelmSy 
Goth, tnam ' maid ' from *ma(g)m, with which compare Goth. 
magu^ ^ hoy' from *mag(fa)iis= At. magAu, The resulting alterna- 
tion of g and tp in different inflections of the same word 
according as the cons, was followed by an t or an « etc, 
was afterwards leveUed in most cases, but has left traces 
in such forms as OE ge^ewen^ gesegen ^seen' (infin. won frt>m 
*seh{w)(m). 

321. h was afterwards weakened to a breath initially, as 
in OE her^ hlud^ and between vowels, as in OHG zehan^ seAan^ 
where it was afterwards dropped: MnG (tseen, zeen), OE 
ten, geieon. A preserved its old sound finally, as in OHG saA^ 
noAl, OE ge^eaAy nieAL 

822. The following is the correspondence of the Gmc to the 
Ar. consonants, starting^ for the sake of convenience, from the 
later Gmc development: — 

h=Ar. * (OE Aeatoan), c (OE feoA, Gmc ^feAu 'property's 
Skjw^w), kA (Gk)th. Adban *have'=Lt Aabere), At^Ax. kU 

j=Ar. j (Goth, juk *yoke'=Sk yugdm). There seems also 
to have been a Gmcjf}', which appears in Goth, as ddj (=ffim?), 
in OIceL as gg; €l^€l^, as in tvaddje^ tv^ggjaszOE ttoagea <of 
two.' 

r=Ar. r (OE r^A^=Lt rectus). 

1= Ar. I (OE lang^Jji longus). 

>=Ar. ^ (Goth. >=Lt tu), tA {OE /Ssfia 'troop,' Gmo 
*fdnfjo compared with Sk pdwtAan ' road '). 

a=Ar. 9 (OE nosu=U ndms). ^«)=Ar. it (OE getous). 

B=: weak Ar. s {*kozandy Lt gustdre). zg^ zd^Ai. zgA^ zdA. 

w=:Ar. «? (OE »a^=Sk vida), weak Ar. k, gA (Goth. «w«#, 
nuwi). There seems also to have been a Gmc vnoy represented 
in Goth, by ggv (=flMi^?), in Olcel. by gg{v)y as in iriggv^ 



92 HISTORY OF ENGLISH SOUNDS. 

trygg = OE treowe * faithful/ Olcel. Iigggva = ^haggwa = OE 
heawan^^hauwan 'hew' (Omc ^hawwan). 

f= Ax. p (OE fader), pk(OEf dm). ftz=:kr.pt. 

11= At. I), n (OE lang, no9u). 

m=Ar. m (OE medu). 

'k.^^kr. g (OE ceosan), j (OE coni=Sk yir«4, OBg ;8;rlf«?). 
^^=weak At. ^«, gn^ ghn. 

g=Ar. ^A (Goth, gast ' stranger '=Lt hostis, OBg ^/^), jh 
(OE gold- OBg zlato\ weak Ar. A (togand), weak Ar. M (OE 
fKi;^^ 'nail ' = Sk nakha). 

t=Ai. d (OE etan). st=Ar. 9t, zd (OE ^teorra, ne^t). hi, ft 
=Ar. ^^, jD^. ^=weak Ar. in, dn, dhn. 

. d=Ar. dk (OE ««^rfi*), weak Ar. t (OE soden), 

p=Ar. i (OE »<«;> *sap' = Sk sabar). ^j»=weak Ar. jmi, fof, 



b=Ar. bk (OE &^ai»), weak Ar. j» {*Aaband), 

828. The change of Qmc ^, ^;y r, z into ^, ef, &, r, as in the 
OHQ participles zogan {zokan\ sodan (sotan), haban {hapan\ 
koran is apparently common West Omc. 

824. Final Omc z is always dropped in WGmc, as in OE 
we (Goth, veis, Gmc ^wlz)^ md *more,' Goth. maM=Gmc *maiz 
(cp Goth, maiza), 

826. Another WGmc change is the doubling of conss. 
before j, which was then dropped in OHG and OE, but kept 
in Old Saxon. Examples are OSaxon mllio (Goth, vilja), 
leggian (Goth, lagjan), biddian (Goth, bidjan), settian (Goth. 
9aijan)y skejppian (Goth, skapjan) = OE toilla, Ifcgan, biddan, 
saltan, 9C(ppan. This doubling is in OSaxon and OE 
confined to conss. which are preceded by a short vowel; 
but the evidence of OHG forms such as leitten from older 
^leitjen (Gmc '^^i^a»)= OSaxon If dan, OE ladan, shows that 
it must once have been carried out, though soon dropped in 
OSaxon and OE. -r/- seems to have developed into rrj in 
OHG (^efrf^= Goth, hau^an 'hear'), but in the other lan- 
guages a parasitic vowel seems to have developed itself 
between the r and the^, which prevented the doubling, as in 



GEBMANIC: HIGH GEBMAN. ^^ 

OE hgre, gen. A^r^tf*=Goth. harji ^Bxmjy A^rian=Qoth. hazjan 
* praise.' 



HIGH GEBMAN CONSONANT-SHIFT. 

826. The second, or High German, consonant-shift is an 
independent, and much later phenomenon. The first shift 
was probably completed (or nearly so) some centuries before 
our era, the second did not begin till probably at least five 
centuries after it. It was a gradual process^ which began in 
the highlands of Southern Germany, being carried out most 
completely in the Alemannic and Bavarian dialects, and 
gradually spread northward to the Frankish dialects — along 
the Rhine even beyond Cologne — resulting in various com- 
promises between High and Low German, included under the 
common name of ^ Middle-German,' to which group Modem 
High German belongs. 

827. The following are the changes which constitute the 
second shift : — 

Gmc kg, xhtd, «J>pbfv 
OHO ch,hh k h z,zz t d ph,ff p v b 

828. The first change was the aspiration of ^, ^, p into H, 
afterwards written cA^ *H, fhy also written gf^ pointing to 
affiication (140), which in the case of z was carried out com- 
pletely in all the dialects. After a vowel all the affricates 
lost their stop, the second element being doubled by way of 
compensation, although the doubling is simplified finally, in 
accordance with the general rule of OHG spelling. Hence 
OE macian, icy ojpen^ Bcip appear in OHG as maAkon, iA, offan^ 
ihf. In one OFrankish text Gmc t between vowels is written 
Z9%y after a vowel at the end of a word xv, as in toazssar^ dAaz%^ 
OE water y pat ; the other OHG mss writing zzy z : wazzar^ 
daz. In MnG this weakened z has become s — vxuser (vasar), 
da89 (das) — but in MHG it never rhymes on *, from which 
it was no doubt distinguished by its dentaUty — Si*. Initially, 
and after conss, and when doubled (both by Gmc and West 
Gmc doubling) k^ t, p remain affricates : kAomy cior»=OE corny 



94 HISTOET OP ENGLISH SOUNDS. 

wercAsztoeorc, we{c)chen ^toecean \ phenning^p^ing^ ielpJian=^ 
helpany 9ke{p)pAen=sc^pan. In the above-mentioned Frankish 
text the strong affiicate of ^ is written z^ that of tt being ex- 
pressed by tz between vowels, by z finally: ri^=OE Ud, iolz:= 
Aolt, 9etzan uos=^ saltan, teaz^scaU. In the other texts they 
write zz in sezzan as in wazzar etc. 

829. In the combination^ sky H, sp, tr, ht^ ft the stops are 
not shifted : /«^=0E Jisc^ 9lein=stdny treian=tredany fehian^ 
feohtan. 

330. The change of ^ etc into k etc, as in kiporan^OE gAorenj 
is confined to a few of the most southern dialects, g etc 
probably had their present South Qm pronunciation Q' etc. 
WQmc gg etc, on the other hand, are regularly unvoiced: 
hrucki = OE hrycgy petti = h^. 

83L Qmc/ was first voiced and then stopped, becoming d^ 
as in dorUy erda-=QiEkJ)omy eorpe. The intermediate w is written 
th and dhy rarely with the OE d. . 

332. Omc/was no doubt voiced like/, but having become 

a lip-teeth instead of a simple lip cons., it was not stopped. 

The resulting > is expressed by / in the oldest texts, in 

accordance with OE usage, afterwards by the Romance u, v : 

fater (vater) =:0'E fader ^ fuot=Jldd, grdvo^gercefa. In some 

late texts v is written initially after a preceding voiced sound, 
/ after a breath : tu vdAest^ ih fdhe^ showing that v really 
meant >. 

333. Gmc r, being a lip-cons., was easily stopped, becoming 
i, as in haban (OE hafen^ Gmc ^kavand) ptc of kef en (Qoth. 
kajjafty OE kfbban) < lift.' This b is sometimes written p in 
the southern dialects. 



fiUNES: GEBMANIO. 95 



EUNES. 



GEBMANIO. 



The consideration of the national Germanic alphabet — 
the Bunic — has been deferred to the present place because its 
development cannot be understood without reference to the 
Qmc consonant-system. 

384. The oldest — common Qmc — Bunic alphabet of 24 
letters is preserved in inscriptions going back to about the 
third century a.d. The order and values of its letters are 
as foUows : 

1^345^78 9 10 II 12 13 14 15 16 

Pnf|^K<XP:Hl-|$1.Wth: 
fu]>arkgw hnijeupzs 

17 18 19 20 21 22 23 24 

t B M M r <> * M 

tbemli) od 

835. As regards the origin of the runes, there can be no 
doubt that they are taken either from the Qreek or the Latin 
alphabet. 

336. The changes of shape in the runes are the result of the 
material on which they were carved, nl slabs of beechwood. 
The first result of this was the substitution of angles for 
curves : whether we derive the runic b and i from the later 
Latin curved or the earlier Latin and Greek angular forms, 
the result would necessarily be the same. So also with «. 
The second main change was the elimination of horizontal 
lines, as shown in the t ; this was to avoid cutting along the 
grain of the wood, which causes spUntering and indistinctness. 
Convenience of cutting led to the substitution of perpendi* 
cular for sloping strokes, as in the a, and also to inversion 
of some of the letters, as we see in the u. So also / would 
have assumed the form it has, whether it was taken from the 
Greek-Latin U or the exclusively Greek T. Finally, we have 
the principle of compactness: the side-strokes are never 
allowed to rise or descend beyond the top or bottom of the 



96 HISTORY OF ENGLISH SOUNDS. 

main stem. Thus, as the side-strokes of F could not be bent 
downwards without confusion with f> a, they had to be 
moved lower down so that they could be turned up with- 
out projecting above the top level of the main stroke. We 
now see that the most refractory letter was E. By turning 
it on its side they got two upright stems, and by simplifying 
the two other strokes into a broken line — ^by which at the 
same time they got rid of the horizontal stroke — ^they evolved 
the runic M • Other changes were made to differentiate letters 
that would otherwise be confounded, as in the case of M ^ 
and M ^9 H >i and t^ n. We see from ^ ^i = Gi'eek-Latin > that 
the main stem might be lengthened. Hence there is no 
necessity for identifying ^ o with the Greek omega ; it may 
just as well be a modification of the angular equivalent of the 
Greek-Latin 0, which would, besides, have been liable to 
confusion with one of the forms of the g-rune ^ which is 
exactly that of an angular O. 

887. The fifteenth rune "Y occurs only in the very oldest 
Scandinavian inscriptions, where it is the regular symbol of 
that front-modified r which stands for Gmc z, as in the nom. 
sg Xf^h'Mi' ff astir =Lt Aostis, original r being represented by 
R, as in H5iR+f^ Aoma. The i' — which has the form ^ (pos- 
sibly only an ornamental variety) in one alphabet — seems to 
poiat to the old Greek-Italian 1 rather than the late Latin Z. 

888. We may now turn to the new letters. ^ i[) is clearly 
a reduplication of < k, and ^j of \i (possibly of <). P w is 
probably a modification of runic H u. 

889. X^ and M ^ appear to be reduplications of < ^ and 
Greek-Latin D respectively. Another view is that they are 
the Greek X ki and ® tA resp. ; but it is difficult to see how 
these letters could have been applied to Gmc sounds which at 
that time were either Ar. ^A, dA or else some modification 
of them — one would rather expect them to be applied to the 
Gmc representatives of Ar. M, A, iA, t. The use of Greek- 
Latin D to express J) seems to point to a stage in which 
^ expressed both t and /, ^ both d and 9 — ^that is, when 
Vomer's Law had begun to work, as is further shown by the 
existence of a runic z. When dA became simple d^ a new sign 



runes: OLD-ENGLISH. 97 

was made by first doubling and then joining back to back two 
^s, so that the old ^ ^ came to have the exclusive value ^, 
which was afterwards made to include that of ^, in order to . 
avoid the greater ambiguity of 1" =both t and /. The value 
of OE 9 was in later times extended in the same way, so 
that it stood for the breath as well as the voiced sound. The 
runic X ^ was probably formed, on the analogy of M, by a 
modification of <, which, in the runic alphabet as in the Latin, 
seems to have been originally the only back cons. 

840. The value of runic J^f does not prove conclusively 
that the runic alphabet must be of Latin origin, for in Greek 
itself such spellings as fHE=W from Ar. ^swe through *9wAe, 
show that the digamma must often have had the sound of o. 

341. The evidence of the forms of the letters is strongly in 
favour of the Latin origin, but chronological arguments show 
that the runes must have been borrowed several centuries 
before our era, at a time when the Qermanic tribes could not 
have been influenced by the Romans, for otherwise sufficient 
time would not be allowed for their divergence from their 
original forms. On the whole the most probable theory 
seems to be that the runes are of indirect Greek origin, and 
that they were adopted by the Goths from some non-Germanic 
tribe of central Russia about the third century b.o. 

OLD-ENGLISH. 

342. The following changes of form and value occur in 
the OE runes. As the soimd a became a very generally in 
OE, F^ naturally took the latter value, a new sign for a 
being formed by a slight modification — ff a. The name of 
F^ was *ansiis, which in OE by regular sound-change becomes 
ds\ hence ^ with a slight modification — f^ — became the symbol 
of 0, Oy keeping the place of the old f^, which with its new 
value was relegated to the end of the alphabet. The name of 
the original o-rune was opil, which in OE became apU ; hence 
St assumed the value os, ce, A sign for y was made by com- 
bining u and i into fr). The diphthong T ea (or eol) is of 
obscure origin. Of the conss. Q J was disused because <D had 

H 



98 HISTORY OF ENGLISH SOUNDS. 

been stopped into m, and as g before front vowels had become 
ffi, the rune x Q^'^ became the symbol of original j as well 
* as of ^. The pure back a was then provided with a special 
symbol by modifying X into M gar. In the OE runes k is 
represented by k cen in various forms, and by %y which is 
evidently a modification of the ^(^r-rune, and probably stands 
for the back a as distinguished from q, of which cen appears 
to be the proper symbol. 

843. The following is the correspondence of the OE and 
Gmc runes to the original Greek or Latin forms. 



Qk-U 


Jiunic 




OE 


OE Names 


F 


V 


f 




feoh 


V 


n 


u 




ur 


I> 


> 


y 




j7om 


A 


f^ 


a 


ae 


sesc 


R 


K 


r 




rSd 


< 


<(OEk)k 


Q 


cen 


«? 


X 


g 


ffl 


gefu 


V? 


l> 


w 




woRn 


H 


H 


h 




ha^l 


N 


+ 


n 




ned 


1 


1 


• 
1 




IS 


II? 


§ 


• 

J 


— 


ger {Gmc ^sftra) 


? 


1 


eu 




eoh (Gmc *ehu) 


? 


M 


P 




peorJ> 


1 


t 


z 




eolh {Gmc *elhaz) 


H.s 


H 


s 




sigel 


T 


t 


t 




tir 


B 


^ 


b 




beorc 


E 


M 


e 




eh? 


M 


M 


m 




mann 


^,'^ 


r 


1 




lagu 


« 


<> 


^g 




ing 





St 





ce 


<pj7el {Gmc *6J?ila) 


I>I>? 


M 


d 




deeg 



OLD-ENGLISH : DIALECTS AND TEXTS. 99 



Additional OE nines: 



K 


a 


ac 


r* 


o 


oe 


T 


y 

ea 


ear 




e 
a 


gar 



OLD-ENGLISH SOUNDS. 

DIALECTS AND TEXTS. 

844. There were four chief dialects of 0£ : (i) Northumbrian 
(North.), {%) Mercian (Merc.), corresponding to the later Mid- 
l«^d» (3) If^est'Saxon (WS), (4) Kentish (Kt). North, and Merc, 
together form the Anglian group. Et represents the dialect 
of , the Jutes, WS that of the Saxons ; together they form the 
Southern group. 

345. The oldest dated ms containing OE words is a Kt 
charter of 679, but some of the Bunic inscriptions are prob- 
ably older. At the end of the 9th cent, a great revival of 
prose Iiteratui:e in WS took place under Alfred, and hence- 
forth WS becomes the official language of the laws and 
charters, although the local dialects are still represented in 
more or less unsophisticated texts. The OE texts up to 9cx> 
preserved in contemporary mss, with the exception of Alfred's 
works and the Chronicle, together with the Runic inscriptions, 
are given in my Oldest English Texts (OET). 

346. North, extended from the Humber to the Forth. 
Early North. (eNorth.) from 700 or earlier to 800 or some- 
what later is scantily represented by Bunic inscriptions, 
short poems and proper names, all printed in OET. Late 
North. (INorth.) of the latter half of the loth cent, is repre- 
sented by the interlinear gloss in the Durham Gospels (Du.) 
and the Durham Bitual (Bit.), both of which are quite free 

H 2 



lOO HISTORY OF ENGLISH SOUNDS. 

both from WS and Scandinavian influence. There are no 
Noiih. characters. 

847. Mercian extended from the Thames northward. 
eMerc. is represented (partly at least) by the Corpus glossary 
(Cp), and in a very pure and consistent form by the Vespasian 
Psalter (VP), which is probably (as also Cp ?) West-Mercian, 
and by some WMerc. charters. IMerc. (South Yorkshire) 
is represented by the interlinear gloss in the Rushworth ms 
of the Gospel of Matthew (Ru.), the gloss on the other gospels 
being a copy of Du. The language of the later Merc, charters 
is greatly mixed with WS. 

348. WS occupied the whole district south of the Thames, 
with the exception of Kent, and apparently of SuiTcy, whose 
dialect seems to be nearly Kt. WS seems also to have 
spread up the valley of the Severn, and so encroached on the 
Merc, dialect. The oldest document of WS is a charter of 
778, followed by one of 847. There are very few eWS 
charters. The most important eWS texts are Alfred's trans- 
lation of Gregory's Pastoral Care (Past.), and of Orosius 
(Or.), together with the Parker text of the Chronicle (PChr), 
all preserved in contemporary mss. IWS is represented most 
purely in -filfric's Homilies (MtcH). 

349. eEt is represented (partly at least) by the Epinal 
and Erfurt glossaries (Ep., Ef.), the former probably written 
at the beginning of the 8th cent., and by numerous charters 
in an apparently pure dialect. lEt is represented by a 
few charters and by glosses on the Proverbs of Solomon 

(Kgl). 

350. Very little is known of the East-Anglian dialects of 
Norfolk and Suffolk, although there are a few late Suffolk 
charters. The East- Anglian dialects (perhaps including that 
of Essex) seem to have had some features in common with 
Kt (which may be partly the result of common WS influence), 
forming with it a special South-Eastem group. 

861. Most of the IWS mss which are copies of earlier ones 
show a mixture of forms of different periods : they never 
retain the older forms consistently, and hardly ever carry out 
their own spellings consistently. There is also a considerable 



OLD-ENGLISH: ORTHOGEAPHT, lOI 

mixture of dialect. This is especially the case in the poetical 
texts, which are mostly loth and nth cent, copies of Anglian 
originals. 

OBTHOGBAPHY. 

852. The Anglo-Saxons brought with them to England 
their national runic alphabet. On their conversion to Chris- 
tianity they adopted the Latin alphabet in its British form. 
At first the Latin and the llunic alphabet continued to be 
used side by side, the one in writing, the other in inscriptions, 
without influencing one another. Li the oldest mss we find 
to expressed by u{u), p by thy and it is not till the 9th century 
that these digraphs are generally superseded by the more 
convenient p and ]? of the Runic alphabet. Some of the in- 
scriptions show a mixture of Roman uncials and runes, 

868. The OE alphabet consisted of the following letters, 
those in () being only occasionally used : a, <«, ^, (^), i, (?, «, y ; 
*, <?, d, ?,/, 9, A, {h), I, m, n,p, (qu), r, *, ^,/, w, x, {z). 

364. Li determining the values of the letters in OE we 
must be guided by the traditional pronunciation of Latin, 
remembering, however, that the pronunciation learnt by the 
Anglo-Saxons was more archaic than that of the Continent. 
The evidence of such Welsh loan-words as ctoyr from Lt cera 
shows that the Celts pronounced Lt (? as Q everywhere, and it 
is to Celtic tradition we must ascribe such OE spellings as 
Cent =^ owe Kent, while in the OHG of the 8th cent, c before 
front vowels was used to express (ts), as in cit=zit z still 
had its original value of (dz), as shown by the spelling ladzaru% 
^Lazarus in Past. OE spelling is also very archaic in its 
retention of y in its original value of Greek u f , while in 
OHQ it had been confounded with j, and almost entirely 
disused. y=f survives to the present day in Swed. and Dan., 
having been introduced into Scandinavia by the Anglo-Saxon 
missionaries. Lastly may be noted the separation of ae^ oe 
and ^, which on the Continent were soon levelled under simple 
e. Here, again, the Scandinavian languages show English 
influence, oe is always written in two letters, but ae is con- 



I02 HISTORY OP ENGLISH SOUNDS. 

tracted into ^, especially in WS, the earlier texts writing in- 
differently acy tB, p (in which the tag is a shortened a), and — 
by omission of the tag of ^ — e. In VP e is regularly written 
for short a^ and ic, ae^ § are reserved for a. In Kt, both early 
and late, e is freely used both for iB and a. In this book the 
§ of the mss is printed <9, to avoid confusion with the normal- 
ized § parallel to g. 

366. There is, however, a good deal of uncertainty about 
OE pronunciation, owing especially to the defects of their 
consonant-symbols, ^, for instance, being a very ambiguous 
letter. Here we must be guided by comparison with the 
cognates, and with the ME and MnE forms, as also by the 
laws and analogies of OE itself, and the variations of its 
spellings. 

STRESS (metre) \ 

866. As quantity and stress are as essential elements of 
metre as time and barring are of music, it follows that the 
metres of a dead language ought to be, or at least may be, 
sure guides to its quantity and stress. 

But in practice it is impossible fully to harmonize the 
natural quantity and stress of a language with the artificial 
quantity and stress of metre : one or other must go to the 
wall. Thus, our present verse is based mainly on the natural 
stress of the language, each strong stress marking the begin- 
ning of a foot (bar). But the stress-groups of ordinary speech 
amount to nothing more than prose: to make these stress- 
groups into metrical feet it is necessary to have them of equal 
(or proportionate) length, and in English verse we lengthen 
or shorten syllables without scruple in order to make the feet 
of the requisite length. In Greek and Latin, on the contrary, 
the language itself supplied the quantities, and the division 
into feet (barring) was effected by an artificial metrical stress 
(ictus), which completely overrode the natural stress of the 
language. It was natural in Greek and Latin to found metre 
on the quantities rather than the stresses of the language 
partly because stress was probably not very prominent, but 

' Rieger : Die alt- und angelsacliBiBche verskunst. 



OLD-ENGLISH: STRESS. IO3 

mainly because of the strictness and clearness of the dis- 
tinctions of quantity and their entire independence of position 
and accent. In English these conditions are so imperfectly 
fulfilled that it is almost impossible to reproduce in it even so 
simple a metre as the Greek hexameter. In some languages, 
especially those which have no marked distinctions of stress 
(and quantity) the natural language supplies nothing but the 
number of syllables, which is strictly adhered to, such a 
variation as a4ia for da (as in the hexameter) not being allowed. 
We have then three varieties of metres, if we class them 
according to their linguistic basis : (1) «^^^-metre, (2) quantity- 
metre, (3) syllable-metre, 

867. There are, of course, endless compromises possible. 
Even in Greek there can be no doubt that the natural quan- 
tities were often forced in metre ; and in English the best 
poets are influenced by an unconscious respect for the natural 
quantities of the language. 

868. Old-English verse is a remarkable instance of such 
a compromise. In it the number of syllables is perfectly 
indifferent, as long as they do not interfere with the other 
conditions. Quantity is rigorously observed within certain 
limits, but the main element is the natural stress of the 
language, both word- and sentence-stress, whose laws are 
observed with great strictness. Alliteration is indissolubly 
connected with stress. Each full (long) verse has four stresses, 
and is divided by the caesura into two half (short) verses, 
bound together by alliteration : one or two accented syllables 
in the first half verse and one in the second beginning with 
any vowels (generally different) or the same consonant, the 
last alliterative syllable being always the last but one : 

)^ac5m inngan | ^aldor ]>egna, 
^^cene monn | domege weor]7od. 

869. The alliterating syllable must not only be the stressed 
one of the word it belongs to, but this word must also have 
the strongest stress of any in the half verse. We know by 
the written accents of OHG mss that in all syntactical com- 
binations of nouns (subet. or adj.) the first member of the 
group had the main stress, the verb being regularly subordinated 



I04 HISTORY OF ENGLISH SOUNDS. 

to the noun, and the metrical laws show the same principle 
was followed in OE also. Thus in such a line as 

^nu Beanstanes | «o]7e gel^ste 
9unu necessarily takes the alliteration because it precedes 
Beanstanes^ and such a line as Beanstanes sunu sope geldste 
would be impossible, because then Beanstanes would take the 
chief stress and alliteration, while, on the other hand, it would 
be quite regular if some verb were substituted for Beanstanes, 
So also in such groups as -^eaden \miere^ -mare ifeoden ^ famous 
prince,' 'iaran :twd, 'twegen vfet^ 'feoden -.Hrofgdr^ 'wordum itcis, 
'Aand and irand 'hand and shield.' Quantitative, half-pro- 
nominal adjectives, such aafela, manig^ call form an exception : 
lealles 'manncynnes. 

860. Pronouns are generally subordinated to verbs as well as 
to nouns, often also to prepositions : nanig heora *pohte^ 'of hiera 
'(Bpelum, Emphatic pronouns, such as self^ opety ale, iegj^er are, 
however, treated like nouns. Even unemphatic personals 
and demonstratives (articles) sometimes take the full stress 
from a noun or another pron. : 'uncer twega * us two,' pu 'me 
sealdesty on '^am doege 'Pisses llfes, 

361. Adverbs are treated like nouns when they form a sort 
of compound with a following noun or verb : 'wide gesiene, 
•feorran cumen^ 'inn gdn, 'hi standan. If the adv. follows the 
verb, it generally loses its stress : 'wlitan furhy 'fehp bper to. 
In other cases adverbs, especially quantitative and intensive (cp 
the corresponding adjj., § 359) do not generally take the stress : 
micle 'leaf re y ful 'picclice, nealles 'swaslice * not plea^santly.' 
Other adverbs, such as hu^ swa^p^^ponne sometimes take the 
stress from the verb, but often not: panon he gersohte^ hu 
'lamp eow'\ 

862. A finite verb is generally subordinated to an in- 
finitive, participle or finite verb dependant on itself, just as if 
they were nouns : ge moton 'gangan, 'bldan wolde^ eoto het 'Sfcgan^ 
her sindon gerf^ede^ cwcedon peet he 'wierek 

868. The fundamental principle of OE sentence-stress 
evidently was to stress the modifying, attributive word, which 
was generally put before the word modified. 



OLD-ENGLTSH : STBESS. IO5 

Composition-stress. 

864. In composition the same principle is evident: the 
modifying word comes first, and takes the stress, as in 'heofon- 
rice, 'hsard-fcg * hard of edge/ 'soj^-fast^ 'eorp-lic^ where the un- 
stressed second element has had its vowel shortened (Qoth. 
-leik). 

Abstract (verbal) substantives compounded with inseparable 
particles throw the stress on to the particle in the same way, 
as in 'and^swarUy and the analogy of Sk makes it probable that 
this was also the parent Ar. accent. The corresponding 
verbs take the stress on the root, the particle being often 
weakened, so that we have in OE such pairs as : 
*and-giet ' intelligence' on'gietan 

•8ef-}?unca * grudge ' of"}?yncan 

•or-}7anc ' device ' a*)?§ncan 

'u)^-gang ' escape ' o^'gangan 

*bi-ggng * worship ' be'gangan 

Substantives corresponding to verbs with separable prefixes 
(361) take the stress on the particle in the same way, thus 
to the separable compound verb 'inn-gdn corresponds the in- 
separable compound subst. 'ifin-gang. 

366. The difierent treatment of substt and verbs is due 
to the fact that andgiet etc were true compounds already in 
parent Gmc, while in on'gietan as well as inn^gdn there was 
only a loose collocation of the two elements, so that the accent 
could be put either on the particle or verb, according as the 
one or the other was more emphatic. In Sk there is no such 
thing as inseparable verb-composition. In a normal Sk 
independent sentence the verb is put at the end and has no 
accent) which is taken by the preceding particle, as in dpa 
gacchati 'he goes away,' while in a dependant sentence the 
particle yields its stress to the verb, as in ^r^ 'pa gacchati * who 
goes away.' The former corresponds to the OE 'inn-gdn, the 
latter to 'wtitan purh^ on'gietan. Such compounds as ongietan 
did not become inseparable till the prefix had lost its in- 
dependent meaning, as also, in many cases, its independence 
of form. 



I06 HISTORY OF ENGLISH SOUNDS. 

366. The same stressing of the root-word is common in 
compaunds of prep. + subst. or adverb, such as td'dag^ ofdune^ 
to'gadre^ be'foran, where we have similar weakenings of the 
first element, qfdune becoming adune, and even dme in dune- 
stigan, be^utan becoming butan, etc. 



QUANTITY. 

Metres 

867. The line quoted above 

)^ com inn-gan | ealdor l^egna 

is an example of the quantitative element in OE verse in its 

simplest form : 

aa aa aa aa 

While the quantity of the unstrest syllables is indifferent, the 
substitution of such a word as cyning for ealdor would spoil the 
verse : it would be too short. 

868. Such a half-verse as cyning on corpre is, on the other 
hand, correct, because am is metrically equivalent to m. 

869. The number of unaccented syllables between the 
stresses may be increased even to three, as in csre%t gesohie, 
S€egde sefe cupe. But such a half- verse as *aresta sohte would 
be impossible, because when a long strest syllable is followed 
by a medial syllable, this medial syllable takes a secondary 
stress, and the verse becomes too heavy. 

370. The following are the five main types of the second 
half-verse, which is more regular in its structure than the 
first (note that the quantity of weak and half-strong syllables, 
and of strong syllables at the end of a verse is indifferent) : 
(a) aa aa : ealdor J^egna 

{b) -a(a) aa a : siJ^J'an serest wear)? 

{c) -a(a) a aa : ond hea]70w&dum 

(d) a a*aa : feond manncynnes 

{e) a-aa a : Su]?d^na folc 

871. It will be observed that the truncation of the second 
foot of (i) and of the first of {c) is made up by the initial weak 

* SieTera : Zur rhythmik des genmuiiBchen allitterationsTerses (Paul and 
Braunes Beitrage, x). 



OLD-ENGLISH: QUANTITY. IO7 

syllables— the ' auftakt,* while (d) and (e) are weighted by the 
additional half-stress, whose quantity is indifferent. Similarly 
in {a) if the second half of the first foot is a half-strest syllable, 
the following stress-syllable may be short, as in 'tcyrd \cft 
ngrej)^ d'a aa being felt as of equal weight with da da, 

872. In {c) the immediate succession of two long stress- 
syllables as in gebun hefdon is generally avoided, either by 
resolving the first stress into two short syllables, as in the 
example above, or else by having the next stress-syllable 
short, as in o» land Bpia. Here there is no compensation else- 
where in the verse. 

878. In the first half- verse the same types re-appear, but 
with certain licenses in the introduction of half-strong and 
weak syllables, which are often associated with double alliter- 
ation. 

Orthography. 

874. The metre enables us to settle the quantity of accented 
vowels with certainty in many cases, but in many it fails. It 
tells us nothing about the quantity of unstrest vowels (for the 
fact that such a word as bindere takes a secondary stress on 
its second syllable has nothing to do with the length of that 
syllable), or of vowels followed by more than one cons. It 
therefore becomes necessary to examine the ms evidence. 

876. Doubling of long vowels is common in the oldest mss, 
and occurs throughout the 0£ period. It is often confined 
to monosyllables, as in oa * ever.' 

876. In VP the short a is written ^, as in cester^ hefde^ bee, 
rk being written ae (a^ §\ as in daely dal, d§lan. The diphthong 
ei having become simple e, this sound is sometimes exprest by 
ei, as in eil^ deid^el, ded in C^^feing Du. In WS ig is used 
to express ^, as in astigge in the Past., wiggend, and in the later 
hig^hl^ etc. 

877. There are a few accents in Cp, and in the loth cent, 
they become common, though there is no ms which accents 
fully. VP has a few doublings, but no accents. Accents and 
doubling are sometimes combined : mif, mtf OET, da BIH. 
Sometimes two accents are written on one vowel (as in a 



I08 HISTORY OF ENGLISH SOUNDS. 

charter of 997). The OE accent is the * apex * of the Latin 
inscriptions. According to the general principles of British 
calligraphy it is generally finished off (like the straight final 
stroke of many letters) with a tag, which has misled most 
German editors into printing it as a circumflex. It is often 
difficult to know what vowel it is put on, but as there can be 
no doubt that it was written upwards, we must assume that 
it is meant to be over the letter where it begins. But it often 
begins distinctly on a preceding or following cons., and is 
sometimes even shifted on to the vowel of another syllable : 
leganurriy hegdnum Leechd., wardn, a/jos Or. In the first instance, 
however, the accentuation may be intentional (cp 381). 

378. As the older editors omit the accents, while Eemble 
and the Germans normalize, and Thorpe sophisticates the ms 
evidence, it is difficult to get at it. Thorpe's Chronicle is, 
however, reliable. So also are the Cambridge ed. of the 
Gospels, Cockayne's Leechdoms, Godwin's GuJ'lac, the Blick- 
ling Homilies (BIH), and my own OET, Past., and Or. 

379r The accuracy of the mss differs greatly. The best 
is the (Lauderdale) Or., which has hardly more than two 
undoubted errors. PChr has nothing doubtful down to 
937 (except weg). Both mss, however, accent sparingly, and 
confine their accents to a few words, such as the pret. /(&•. 
The Cambridge ms of iEfcH printed by Thorpe (but not as 
printed by him) is fairly good, as also the WS Gospel of Mtt. 
The Past, accents freely, but often very inaccurately : it would 
be easy to prove from this text that every vowel in the 
language was long ; if, however, we disregard every case in 
which a word is accented only once, most of the anomalies 
disappear. In the following details a single occurrence of a 
doubled vowel, or of an accent in Or. has been considered 
authoritative, but, as a rule, no quotations have been made 
from the accented words of the other texts unless the word (or 
some inflection etc of it) occurs at least twice in the same ms. 
Wherever an isolated form is quoted from these texts, it is 
enclosed in (). 

880. In some cases it seems doubtful whether the accent 
was not meant to indicate something else than quantity. 



OLD-ENGLISH I QUANTITY. IO9 

neopauard^ ediaelle in Cp, together with uuiltmt (OET), seem to 
show that ff , uu were used for consonantal u, uu=to. Cp {tis = Lt 
fus in the Leechd. Such accentuations as fdtu (twice on one 
page in Past.), Spene (Past.), cyning (Chr), gecuran (Chr), ahredde 
(^fcH) cannot possibly indicate length, which would be against 
metre and the whole history of the language, and if they mean 
anything at all, it must be stress, which the scribe confused 
with quantity. 

881. But iiiere are many accents which cannot be anything 
but the result of pure carelessness. The accents not being 
required by the reader (I myself being able to read an un- 
accented quite as fluently as an accented text), came to be 
regarded as ornaments, without which the page had a bare 
look, and were consequently partly written mechanically, 
partly dashed in almost at random. Sometimes, of course, 
quantity-marks are a help, as in the case of GodejiA. gdd, which 
latter is often written good, g6d in the homilies, the striving 
after distinction being evident in such a collocation as godei 
good in BIH. Hence when a scribe deliberately writes God 
with an accent, as happens once in the Or. and several times in 
^fcH, we can only ascribe it to careless neglect of the context. 
So also when we find in GuJ^l. ie write subj. pret. followed by 
the correct pres. ic write. A very puzzling feature of some 
later texts, such as two mss of the Leechd., is their accentua- 
tion of inflectional syllables : hog&9^ wardn, buterdn, namdn^ 
iylfduy drincdn, gehwadum^ langum, wearmum^ wundum, nemne^. 
But this is probably merely the result of dashing in the 
accents after the page has been written, the accent being 
meant for the preceding syllable. It often happens that the 
accents get worse in the middle of a ms. Thus in Past, they 
seem more careless after pp. 70-80, and in the WS Luke there 
is a marked change for the worse after cap. 1 2. 

882. The results of metrical and accentual evidence can 
often be confirmed by that of ME, especially as shown in 
Orm's spelling, and MnE. 

888. The lengthening of final strest vowels is proved by 
the accents : «/, ge/, H Du. ; hi Ru. ; ^/, hi MioBL ; ^, //, w/, 
/», hi^ hig Mtt. In Mtt the emphatic pronominal se is often 



no HISTORY OF ENGLISH SOUNDS. 

accented : J^is ys se he pam pe gecweden y%, . ; se 9e towyrpS, . ; ie 
hyd. . ; %6 pe S€g9. The article se is not accented in Jiftt, and 
rarely elsewhere. There is no accentual authority for a dis- 
tinction between ne ' neither ' and ne ' not/ nS occurring only 
sporadically in both senses — ofbenest in that of * not.' 

884. There is unmistakable evidence of lengthening before 
single conss. in subordinate monosyllabic words : 

6f (prep, and adv.) Past., Or., Chr. 

6n (prep, and adv.) Past., Or., Du., iEfcH. Also dn in Or. 

dc (conj.) BIH, Leechd. Orm has ace, 

at prp. VP. i^at VP. giet Past, gyt, git MicB. ; gyt often 
elsewhere. 

For other words the evidence is not so full : 

is (OET), Past (often), and elsewhere, hts Chr. fi^is Chr. 
Pronominal p^9 in Mtt. 

Mm Or. 

ie Ct. mec Du. 

hU (OET). hwtit iEfcH. 

886. Lengthening in the following prefixes seems to fall 
under the same head : 

Sr^wene .£fcH, Srsargnesse Past. 

dn-syn Mtt. 

un-asecgendlicy ungemetlicy unmiltsting Or. tin- Du., iEfcH, Mtt, 

GuJ^l. 

wdnn-spedig Or.^wa^-. 
fdrdSn in Or. is more doubtful. 

386. If the lengthenings of en and of un- etc are parallel, 
we should expect to find the prefix a- (which was probably 
short at an earlier period) lengthened in the same way as 9vi>d 
etc. That this was the case is proved by the accents : drede 
etc OET, d%endan etc Or., Past., Ru., iEfcH, GuJ^l. (not'in Du. ?). 
The later weakenings oion and of seem to have been lengthened 
by the analogy of this older a-: {dweg^ dmang^ ddune)=ottw^y 
ongetnang^ cfdune Chr. fMrac PChr shows that the older prefix 
tu was levelled under the prp. to, 

387. There are similar lengthenings in strong words also : 
boor^ goor Cp. foor (?) Ep. wer * man * Past. 

wU Past. 



OLD-ENGLISH : QUANTITY. 1 1 1 

paat^pa^ Cp. 

loob Cp, Id/Past, (often), iEfcH, and elsewhere, loffin O. 
fraam, haam Ep. 
bSt{t) compar., IStwrencas^ andfftt(e) Past. 

888. These lengthenings seem to be the result of analogy: 
fffr-, for instance, becoming Hn- because there are several 
common words ending in ^un {dun, tun etc), but only rare and 
doubtful ones in -tin {gepun, gestun, gemun). on, again, is the 
only word in -o», except /(?» and hwan, which are practically 
mere variants oifam and hwam, and are frequently written 
Pan, hwan (there are, of course, many words in -p»=-a», such 
as gemgn, swgn). toer and wel are almost the only words in -el 
and -er (except the imperatives her, stel, which would follow 
the analogy of beran, stelan), there being also very few in -^, 
and none in -fl. The only words in -of are of, kof, hf, while 
there are many in o {hqf sb, prt, behaf, hrof, glqf etc). The only 
words in -or are bar, gor, spar, dor, the last two being probably 
long as well as the others, get ^ yet ' is almost the only word 
in -et besides (and)get and set. The only word in ^t is b(t, 
where the analogy of meaning of betan may have helped the 
lengthening. Short monosyllables seem to be altogether in 
a minority in OE. 

889. The original short vowels were no doubt preserved (i) 
in inflections: Idf, hfes, lofw, (2) when the word was un- 
emphatie. 

890. 9ooe in Ep. for the normal 9occ of Cp and buuc in Cp 
^bucc 'buccula' of Ep. look like a confusion between conso- 
nant- and vowel-length. So also ic Ann Ct, m6n, mSnn, mdn, 
mdnn, mSn, Past., mdn Or., mdn, mdn Chr. Many of these last 
are the impersonal man * one,' which was weak, and probably 
had a short n, and therefore falls under the same head as on. 
he ong{t{t) Past. {94^, 9iss) Past. 

891. (ar^Mos, wdec, dcce, nnrddt^unrot, bidtende:=bitende) in 
Du. show the confusion still more clearly : here the quantity 
is indicated twice over. In (Hnin, Mnnen) the later history of 
the words seems to show that there was real doubling of the 
cons., as in dttor etc below. 

892. Accentuation of short syllables followed by an un- 



H2 HISTORY OP ENGLISH SOUNDS. 

strest syllable is mostly sporadic (380). But not in the 
following cases : 

herian iEfcH. 

Mne ^fcH often. Analogy of M% 

hafenkasty hdfene ^fcH. 6fen Mtt. dfer Past, (cp %nAer 
Past.). 

(iSsad, gelSms) Du. 

wMere John ; analogy of wAfl or confusion with *ioaUre ? 
ic Ste {dtafiy Stad^ St) Mark, and similarly in Du. 

fdder GuJ^l. 

393. w lengthens in the WS -Iw : Mw Mta, niwe iEfc, Mtt, 
Luke. 

394. g seems to have a lengthening influence : 
d(^g iEfcH, Luke and elsewhere. 

toSg Past., very often. 

fiSga^ hdgian ^fcH. 

gedegen^ hrdgel GuJ>l. 

Scgum Chr. 
So also in the verbal endings: faHnaagiS Du., wuntggendum 
BIH (gelacnigan), geendigSan^ manianne {^ondai^) Past. Here, 
however, the lengthening may be partly compensatory. The 
lengthening in dag etc seems to be explained by iEfc's 
spellings daig^ weig^ the accent really indicating the glide: 

396. The lengthenings before vowellikes + cons. (' group- 
lengthenings ') are important : 

r : rj> : f6r]) GuJ^l. 

m : dm -^fcH, Mark, Luke. herSm Du. yman Leechd. 
h&nie Leechd. ^m Ct, Gu)?l. 

rt : Sart BIH. Pdrt Chr. 

rd : Iieardige Leechd. ^geard Chr. f^rdy fird Chr. wurdan 
Chr. 6rde Verc. w6rd iEfcH, Du., Luke. getcSrden Luke, hroord 
Ef is probably an error of the German scribe, {wyrd^ forwyrd) 
BIH. 

1 : 11 : walUy fuel [not in heUe^ sellan, etc.] VP. aalle (dlle) 
Du. 

If: cal/Yf. 

Im : cwalman VP. 



OLD-ENGLISH : QUANTITY. II3 

It : gemaltan VP. 

Id : €Bldo, onAaldan VP. haaldum = ealdum, did, sdld, Du. gdld Bu. 

n: nn: ofemodnn, angdnn MiaS.. [But cp § 390.] 

no : {stennc, sUnco) Du. tost^ncte ^fcH. inc Mark, drinean 
Leechd. 

ng: (onfeinffon, gefeng) Du. \feng very firequent in Or.; 
never accented.] sprdng, gestrdngian .^cH. €Bi^el only once in 
VP ; engel frequent in VP. 

nd : dnda ^fcH ; Adnd Or., Luke ; hSnd, Idnd Du. ; bewdnd, 
fdndian iEfcH. synd Luke ; ^dnd Or., Luke, BIH. Snde Du., Luke ; 
geendod ^fcH; sindan Mtt. Mnd * hundred' Du. ; gehunden 
^fcH ; apunden Leechd. ; Lundenbyrig Chr. ; gem^nd elsewhere. 

mp : beldmp .£fcH ; geldmp Qu)'l. 

mb : Idmb iSfcH, John. 

396. We may sum up by saying that these lengthenings were 
absent from the earliest WS, but fully developed in later WS and 
Anglian, those before U being early Mercian as well in most cases. 

397. Long vowels are kept before so and at : 

gaast, gdst Du., gdst iEfcH. mdst Or., maaste Du. mast^ 
driB%tan VP. weste{ne) Mark, diist Leechd. Hence lengthening 
in Crist iEfcH. 

398. Lastly, we have isolated lengthenings before 
gn: ^e Du. 

fh : st^ne (slefen), Luke. 

It: 4ft Ghr. 

fd : Ju^de GuJ^l. 

z : fdxes Leechd. betwux .^cH. 

399. Length seems to have been kept in such words as dttor, 
to judge from [gedtirap, g6ddra, gdddre, hluddre) in iEfcH. up Past, 
upahebban Or., appears as dpp iBfcH, tip{p) Mtt, the p being 
doubled by the analogy of uppan. The older form was no 
doubt up, the later upp, the spelling upp being a confusion. 

400. Compensatory lengthening is seen in sw^ra Leechd., 

ftra Verc, hool Ep,:=iolA, hdles Ct; and where an h has been 

dropped after the r and /, it is confirmed by the metre^ which, 

however, shows that the lengthening was constant only in 

those words in which the original h was followed by a vowel : 

I 



114 HISTOBT OP ENGLISH SOUNDS. 

fvran {firiho gpl OSaxon), smora^ anettan=.*onhai^an^ dretta=: 
*vrAatt}a. The short vowel in such a form as feore*^ which 
the metre shows to have existed side by side with the normal 
feorety is explained by the influence of the uninflected feorh. 
The metre also shows pytelypyrel^ due to an older alternation 
^Pyrel^ *Jiyrles from *Pfrhily *pyrhle9. 

401. The dropping of g (chiefly in IWS) before /, «, d 
lengthens the preceding vowel : 

ill Past. 

rh, Or., iSfcH. gAi Verc. angSn Mtt befrCnende^ hefrin ^fcH. 
/^ Mtt, Luke. 

geiM^ heUd Or. %&de^ lede JSfcH, Luke, brdd Qu]?l. 

402. When, in inflection, derivation, or composition, another 
cons, is added to a cons, preceded by a long vowel, the length 
of the vowel seems to be generally kept : 

heriumian Du. rixian Chr. giU%ing Past. 

to dSnne Past., iEfcH. 

wifman Or. 

wingeard Mtt. 

hiit {^=iaUy idtej))y scft {=:sceotefi) Or. IdH (^^ladep, laiefi) 
MioB., gemeUe'ULii, ftettNV, A^deDn. l^dejf^rdeLvike. ge- 
heende Cp. geflymde Chr. geScde Du, afSdde Or. hBdde VP. 
lAdde Gu];l. d6ndum BIBL vmdom Or., Past., Luke. 

408. Shortening before cons.-groups is uncertain, the evi- 
dence being, of course, purely negative. I have not met 
with any accents before hi in such words as hrohie^ sohte^ ge- 
fohty nor in the forms minne, Jinne from mln^^m, VP, however, 
has fueht^ nuBhtun. 

404. Weakening of stress in composition only occasionally 
shortens, -dom and -had seem to have always kept their long 
vowels : -^Sm Or., Verc, -rfo(w» Past. ; -had Past., -^fcH. -lie and 
-red (in proper names ; cp § 450) are very rarely accented. 
The shortness of -lie is proved by the frequent -/<?c, and by the 
metre, which, however, often shows i, especially in the adver- 
bial "Rce. The pronoun Us is accented in Du., Past., and iEfcH. 

406. The vowels of inflectional and derivative syllables 
appear to have been shortened throughout, as shown by their 
interchange with one another and weakening into e. 



OLI>-ENGLISH : QUANTITY. II5 

406. Lengtheniiig of (b in strong preterites : hrdcy wrde^ sdt^ 
b<6d Or., 9^t Chr., geb(h iBfcH, together with gendm ^fcH, seems 
to be due to the influence of the plurals hrdcouy ndmon etc. toaa 
waa probably protected from this lengthening by its want 
of stress and difference of cons, (pi wiiranjy but Chr. has tviis, 
^tt, ^t in Du. shows original lengthening (Olcel. at), 

407. The accentuation of the diphthongs calls for some 
remarks. In Or., Du., iSfcH ea^ eo are often accented, ea^ eo 
hardly ever, but other mss confuse them more or less. On 
the whole, the intention seems to be to put the accent on the 
first element : ^a, h. In iBfcH both elements are sometimes 
accented : ed^ U6f, In Or. ie ' river ' alternates with i/, and in 
iEfcH iu « formerly ' with ii. 

408. Foreign words appear to have had their strest vowels 
long, the metre showing Saian^ Adam, Eve, Isac, David, Maria 
etc. In iBfcH, however, we find Adam, Maria, BIH has 
Addm, In these words every prominent syllable was prob- 
ably strest pretty evenly. 

Consonant-quantity. 

409. In 0£ metre such a word as tvifme is exactly equiva- 
lent to wine, wine ^ friend ' being equivalent only to a mono- 
syllable such as win or wynn. Again, two such words as in 
the prp and inn the sb and adv are kept strictly apart by 
the double n ; for, although inn is often written in, the n of 
the prp in is never doubled, except in very late mss verging 
on ME, where the distinction of quantity in final conss. was 
lost Finally, of course, the distinction could only have 
been purely quantitative : ll, tl*, but between vowels it was 
probably syllabic, the second cons, beginning a new syllable : 
8ll[, sIi-tJt, as in the MnE pen-inife ; such, at least, is the 
pronunciation of Mnlcel., Swedish and Italian. In eWS final 
cons, length is preserved after long as well as short vowels, 
as in bebiett 'commands ' ; but in IWS it seems to be jshortened 
in such cases : bebyL IWS also shows a tendency to shorten 
conss. medially in unstrest syllables, as in the gen. veafene^, 
gen. fem. oj^e ==eWS westennes, qperre. Shortening of doubling 
before another cons., as in the ace. masc. ealne^ midne^eallne^ 

I 2 



Il6 HISTORY OF ENGLISH SOUNDS. 

middne, may be only a graphic abbreviation, as it certainly 
is finally in eal^ bed eio= eall^ b^dd, geiL ealles^ h^des. The 
doubling of g is usually 'written eg (5ji), and this is scarcely 
ever shortened. 

410. OE cons.-doubling is partly Ar. and Gmc (316), 
partly WGmc (325), partly the result of special OE develop- 
ments. The oldest of these last is the doubling of the breath- 
stops c^ t, p together with A before r and /, alternating how- 
ever with retention of the single cons.: bit{t)er, appel from 
Gmc ^bitra (Goth baitr)^ ^apla, Noi'th. taAAer (=WS tear) 
from Gmc ^tahra. The variation between double and single 
conss. not only in different dialects but in the same text 
seems to point to an original inflection *bit{o)r (cp otr Ep.= 
the later otor^ ottor in Cp), biUres etc, the development of the 
parasite-vowel preventing the doubling in the uninfiected form. 
Doubling of t and d after long vowels, as in dttor (air in Ep.) 
•poison*, nceddre^foddoT seems to be later. In IWS doubling 
after a long vowel before the r of the compar. is very com-i 
mon : riccra, deoppra^ whence, by analogy, the adv deoppor. 

411. Other doublings are the result of syncopation and 
assimilation, as in ladde, latt prt and 3 prs of iSdan, from 
*ladede, lade]), 

VOWELS. 

412. The following is the OE vowel-system in its normal 
e WS form : 

a 9 ffi i e u o 

ea eo 

g ie y ce 

a a 1 e u o 

ea eo 

ie ? CB 



J J 



J» 



X 


X 


[ 


i 


} 


Jl 




[» 






C 


I 




f 


{ 


X* 


It 


[♦ 


}« 


}♦ 


XW 




[» 







!♦ £♦ {♦ 



OLD-ENGLISH : VOWELS. 1 1 7 

The chaDges by which the OE vowel-system developed out 
of the Gmc are partly isolative, but mainly combinative. 

a, 89, 9 

413. The most important of the isolative changes is that 
of Gmc a j into a x, which is common OE iii such words as 
tr«*, €ecer^ fader ^Qo^. vas^ atr^fadar. When an a, or u 
follows, WS generally has a, thus daff has pi da^^as, dat. pi 
dagum. So also in hafoc ^ hawk '. But in the oldest texts we 
find such forms as habuc (Ef.), hebuc (Ep.), where tf=x (354)> 
the later Cp showing habuc. So also Ep. has besu and baso 
against the baso of the Leiden gl. and WS. Ef., again, has 
Aara against the kara of Ep. and Cp. 

414. As we see, a is sometimes written e in the oldes 
texts. This spelling is regular in VP, where a is kept for a ; 
it is common also in Et. 

416. Gmc a does not change to a before nasals. Ep. 
always writes a : gimang^ ganoty hand^ 9camu. Cp sometimes 
writes a, but generally 0: onga, Aond, scomo. VP writes 
invariably. eNorth. generally and INorth. always write 0^ 
a appearing only in the oldest texts. eWS writes a and 
promiscuously, although the rarer words seem to favour a, 
the commoner 0. IWS writes a only. Early and late Et 
write a and promiscuously. We may sum up by saying 
that the Anglian dialects favour o^ the Southern a. The a 
before nasals was at first no doubt simply jf, which was 
afterwards rounded, the nasality being gradually lost, giving 
J — original Gmc 0, as in eom^ being 3 — j^st as Gmc j*J 
became OE ^(458). It is possible that the fluctuation be- 
tween a and in the earlier period is purely graphic, j lying 
between j and }, and therefore capable of being expre49sed 
either by a or o. 

416. In unemphatic words, such as on^Jxmey Jumne^ the is 
regularly preserved even in IWS, which also separates the 
unstrest otin in onginnan from the strest an- in anginn, eWS 
pnginn. In these unemphatic words the g no doubt became 
(?=}, which in IMerc. sometimes becomes «, as in ollung= 
gndlgng. 



Il8 HISTOET OF ENGLISH SOUNDS. 

417. In lEt the o o{ on etc becomeB a : an, fane^ panne^ 
anginnan. So also the -on of inflections becomes -a» in IKt, 
as in w^ran:=zwieron, earlier toarun, 

418. Unstrest Gmc a often becomes o under the influence 
of a lip cons. Regularly in ^ as a prp and unstrest prefix, 
contrasting with the strest of- in irfjmnca (364). w has the 
same influence in erfe^word (VP), to-word (VP)=WS ieffe- 
weard^ (dweard, WS hlaford from *hldf'W(e)ard, WS Os-icold, 6 
in Grim-boldy p in herpop^^h^e-pa^ * troop-path '. This is 
often -vreakened to a (cp § 417): herpap^ hlafard (VP). The 
occasional eWS was^=.w(Bs seems to point to *«?o*. 

419. As the change of a to x is carried out in Frisian — 
■where however the \ is written tf, as in VP and K1>— except 
before nasals, we may assume that Gmc a became (B every- 
where in Anglo-Frisian, except when nasalized. 

420. The apparently anomalous vowels in the preterites 
am^ barn, harst and in gcers are explained by the r-shifting 
(510): the earlier forms of these words were ^rann, ^brann, 
^brast, ffras, which last is still preserved in the oldest texts. 



421. OE i is generally Gmc i, as in mtan, bindan. Gmc 
e becomes i in OE before nasals, as in niman, OHG neman, 
OIceL nema. As ie had the sound of X in WS (474), i must 
have had that of the narrow X. 



422. Answers to Gmc e: beran, neH. As § had the open 
sound (468), e probably had the close sound [. 



u 



423. Answers to Gmc u, as in mnu, gebunden, and to Gmc 
before nasals, as in genumen=^OH.Q ginoman, punor=^OS.Q 
donar (cp § 4Si)- Gmc also becomes u in OE after or 
before a lip cons., as in wulfffull, ufan, bucca^OH.Q wolf, vol, 
obana, boccA, The analogy of i makes it probable that the 
OE u had the narrow sound i. 



OLD-ENGLISH: VOWELS. IIQ 

424. u preceded by te; is often the result of the influence of 

that cons, on a following to from i (431, 434), as in toudu=s 

older toiodu, totdu. So also in widutoe (widwe YP), louee < week.' 

wu' from weo^ is frequent in IWS : swurd, €WUiter=eWS noeard^ 

^weosior. 

o 

426. Answers to Qmc 0, as in gecoren^ gold, god. As g had 
the open sound, probably had the close sound }. 

426. preceded by w is often a later smoothing of eo, as in 
WS woruld from weoruld (VP), INorth. noord from sweord. 

ea, eo 

427. These diphthongs are mainly the result of parasiting 
(159). Qmc e before r + cons. becomes eo, sometimes written 
to: steorra, eorfie, iorpe^OSsx.on sterro, eriia. The undiph- 
thonged forms are still occasionally preserved in the oldest 
texts : hertk, smerwi in Ep., Bemhard in a very old Et charter. 
Gmc a becomes ea under the same conditions : earm, heard=: 
OSaxon arm, hard. In the oldest texts we sometimes find this 
ea written eo: weorrae, seonoum (Cp), Georored, TJulfheord (LV). 
It seems probable, therefore, that the voice-glide between the 
vowel and the r developed into a full glide-0 in both diph- 
thongs. The later divergence is the result of the difference 
in the first element. In ^(7 = [i- the was supported by the 
close e, while in *haord from AFrisian *hard it was first 
broadened to j by the influence of the preceding \, and then 
unrounded, exactly as in the long ea (459). It is probable 
that the first element of ea was always a, e being written 
for convenience. Traces of the original AFrisian forms 
seem to be preserved in such spellings as geruua in Ep., 
Bemhaerd in one of the oldest Kt charters, where we also 
find hoard, 

428. In the Southern dialects Gmc a becomes ea before 
/+cons. : feallan, ^a/tfs OSaxon fallan, aid. AFrisian a in 
JEldred (Et charter), wall (WS charter of 847). e generally 
remains unchanged before /+cons. : stcellan, helm. For the 
apparent exceptions see § 433. 

428. Gmc a becomes ea before A=c, that is, h not followed 



I20 HISTOEY OF ENGLISH SOUNDS. 

by a vowel (50!^) in WS and Kt: geseah^ eahta, weax (sf=As), 
The parallel eo from Gmc e is constant in WS in seoA imper,, 
feoAtan, but such forms as reoAt, cneoAt occur only sporadically 
in the earliest WS and Kt, which probably had seox *six* 
also. Original i becomes eo (io) in the same way : meox * dung' 
from Gmc ^miAstu, PeoAtas * Picts '. 

430. Another source of ea, eo is the development of a front 
glide after the front conss c, g (535) in WS, where the c^-, 
ga^ (from Gmc ia^, ga-) of the other dialects appear as c<?a-, 
ffea-, as in ceaster^ sceafl, ffeaf^noJiWS cester (VP), scapt (Ep.), 
gcrf. So also go-, gu- (from Gmc ^o-, ju-) become geo- in WS, 
as in geoc^ geong^ also written gwc^ giong^ giung in eWS. The 
other dialects write geoc (as in Du.), gioc, ioc (as in Ru.), giung, 
gung, iung (VP). The analogy of gcef makes it probable that the 
non-WS ^^-, gir- mean the same thing as the simple g or i, nl m. 
The WS change of the u of giung into shows, on the other 
hand, that in WS the glide after the g must have developed 
into a full vowel, capable of forming the first element of a 
diphthong, which at first must have had the stress on its 
second element, the stress being afterwards thrown back by 
the analogy of the other eas and eo&. This applies also to geaf, 
which must have passed through the stages gaf^ SJ^A 9^^y 
gSaf. 

431. The next main source of these diphthongs is the in- 
fluence of a following back vowel on AFrisian (b and on e, e^ 
which is most consistently carried out in VP: fet (=iftBt), 
pi featu, gen. pi feata, fearan^ scijp, pi sceopu, nioman, beorUy 
pi heoraJf, subj. pres. ic here. These diphthongings are later 
than those which are due to cons, influence. Ep. still keeps 
the unmodified vowels in such words as teru, gelu^ stela against 
the teoru, geolu, ateola of the later Cp, geolu occurring only 
once in Ep. The development of the parasite-^a seems to 
have been quite parallel to that of the other ea, as shown in 
0uch spellings as beom in Cp=the bem (^basu) of Ep. It 
is doubtful whether the influence of a is not really due to an 
older : beoran from *beron. This is probably the case with 
such words as ateola at least, for such nominatives as bogo^ 
later boga are preserved even in Cp. 



OLD-ENGLISH: VOWELS, 121 

482. In WS the eo in such words as heorot^ kerfon is generally 
preserved, but the simple vowel is often restored, especially 
in IWS, the alternation of e-eo^ i-io (eo) in inflection being 
especially avoided, except in the earliest texts ; hence scip 
generally has pi scipu^ etc. The ea is generally eliminated 
altogether, except in a few words, in most of which, such as 
nearuyfealu, it may be really due to the inflected forms neance, 
fealwe^ where it is parallel to the ea of heard etc. ealu (gen. 
eahp) probably owes its ea to the analogy of nearu etc. For 
ceam etc see § 535. 

433. The diphthonging of e before /+cons. seems to be 
generally the result of the development of a parasite u (or 6) 
between the / and the cons., which it then diphthonged the 
e. This is certainly the case with heolstor ^hiding-place' 
from Gmc *kelstro, for the intermediate forms kelostr^ helustraa 
are preserved in Ep., and probably with seolh * seal ', and the 
Anglian seolf=^ eWS self. The development of seolfor and 
meolc out of Gmc ^silovra and *miluk is parallel, except that 
here the vowels after the I are original, not parasitic. 

434. The diphthongings of i and e are regularly distin- 
guished as io and eo resp. in Cp and the other early texts. 
Thus Cp has iioludunj liojmwacy sionu^ suiopan on the one 
hand, and toeorod, beorende^ feotur on the other, =WS tiledon^ 
lijmwdc^ sinuy staipan; fcerod, berende, fetor. But io is written 
occasionally not only for the vowel-diphthonged eo, as in 
scriopu (Cp)=WS screpe, but also for the consonant-diph- 
thonged eo, as in iorpan^ wiorfi, Biam- (OET). Afterwards 
€0 supplants io everywhere in WS. The form Aieora=Ai(o)ra 
occurs once in Or. and in a later Merc, charter, and is evidently 
the forerunner of the late Ayora etc, which are apparently 
Kt. These spellings point to X as the first element of the 
vowel-diphthonged io. 

436. The second element o{ eo,io ia very rarely written n, 
as in Triumuini (BH), Friujmulf (LV )• More important is its 
weakening into a : 

While WS has eoy io m eom^ hiara, heora and other sub- 
ordinate words, yP has ea in earn, heara and in jkara (from 
*^^o, Gmc *Jnz5\ to the exclusion of eo^ except that heora 



122 HISTORY OP ENGLISH SOUNDS. 

occurs twice. So also Du. haa both iiora and Aiara^ while 
Ru. has only Mora, Aeora. Even in WS earn occurs once in 
Or., and ea is fixed in eart. The original relation no doubt 
was that these words had two forms, one emphatic (strong) 
with eoy the other unemphatic (weak) with ea, WS tending 
to generalize the strong, VP the weak forms. The first effect 
of want of stress was to diminish the rounding of the o, and 
then to unround it completely, as in the development of ea 
(459). ea for eo in strong words is only a sporadic irregu- 
larity in VP, but in INorth. it is very common, especially in 
the vowel-diphthongs: weala, eata^YF weola, eotan, WS wela, 
etan. The change of toeola into iceala is probably due to the 
analogy of Aeora : Aeara etc. Even WS has feala by the side 
of fela, older fe(o)lUy the first being perhaps originally a weak 
form of the two latter. The spelling ia for to, eo is specially 
eKt: wiaAt; wiarald,fiah. 

436. The relation of ^a to a is, to a great extent, parallel 
to that between eo and ea. In the eEt charters Aeard etc 
is the regular form when the word is the first part of . 
a compound name, Aard when it is the second (unstrest) 
element : ChidAard, BemAard. So also the occasional eWS was 
appears to be the unemphatic form of v)€es — if it is not a weak- 
ening of ^wo8 (418). The a of Aard cannot be anything but a 
modification of AFrisian *Aard, and BemAard actually occurs 
in one of the eEt charters. The influence of the r on strest 
a was confined to the development of a parasitic a (or g Yj 
after it, while r was able to change unstrest a completely 
into a. a for ea in strest syllables appears sporadically in 
Cp: bisparrade, sarwo, and elsewhere, being especially com- 
mon in eNorth. ; thus CH has hamum, uard, but no ea. The 
later LV, on the other hand, has regularly ea, a for ea 
before /+cons. is common in eWS and eEt, and universal 
in Anglian, where it was probably long (395) : all, Aalm, aid, 
Ai^=:eWS, eEt {e)all etc. IWS has only eall etc, and in 
lEt ea is almost universal. The most probable explanation 
of the Anglian Aard, aU is that they are extensions of 
originally weak variations of Aeard, eall, 

437. The relation of ea to a in vowel-diphthongings is analo- 



OLD-ENGLISH: VOWELS. 1 23 

gons to that in conBonant-diphthongings. Just as VP has heard 
but ally BO also it opposes an a in galan, ivyrtwalan etc to the 
ea in fearan etc. So also Cp has wagnfeam (but weak a in 
€Bafaru)y onseacan, geweada pi, by the side of hara^ gelapade^ but 
only 9calu, stalu pi. The other dialects all show ea in their 
early stages side by side with a, except that in Ep. there is no 
ea except in eceaba etc (535)* But in the Kt charters we find 
such forms as pealntl^ icjkajie, and in eNorth. Eafa, Eafu^ Eadu, 
In WS we find (with a few doubtful exceptions, and after ec 
etc) only a : fatu^fata^faran, as also in INorth. : fara etc. 

438. The variations between ea and a (a), eo and e hitherto 
considered are due to independent divergencies, but there are 
also cases of direct smoothing of ea and eo (cp the similar treat- 
ment of ea and ^(7, § 462,465) due to the influence of a following 
c or ^, and which may therefore be described as 'c-smoothings.' 
They form the most marked characteristics of the Anglian 
dialects. In YP and North, ea before i, x, ht becomes m 
(which in W^d)\ gewehy waxan^ gepceht, Cp also has ^, as 
in io€BxU, laxy against the leax of Ep., which, however, also 
has a in ax, actath. This ts of Ep. is probably the original 
AFrisian ^, which makes it possible that the Anglian a=ea 
may be really original JEhcha in a very early Kt charter is no 
doubt original ; later Kt has ea^ as in WS. <?-smoothing can 
also be transmitted through a preceding vowellike cons. : ea 
before re, rg {rh) becomes e {z=e or a) in VP, e or a in North. : 
ere, (BTCy berg=:W8 earc *ark', bearg *hog'. Cp generally has 
19, as in sparca, mar A ^ horse', * marrow ', rarely ea, which is 
general in Ep., although Ep. too has a, but only before rA : 
farAf marAy but mearc, bearug, which last is also the form in Cp. 

488. eo becomes e in Anglian in the same way before A and 
re, rg, rA: feA, feAtan, were, berg, berAt=iW8 feoA, feoAtan, weore, 
beorg, beorAt So also regularly not only in Cp, but also in 
Ep. : Porgifect, uuerc, duerg. 

440. c and g also have an influence in preventing or 
smoothing vowel-diphthonging. In VP they generate a pre- 
ceding a: dagaa, dagum, cwacian. North, a^ees with WS 
and Kt in having a in these forms: dagas etc. VP itself 
has occasionally a, Cp has a, a, but generally ea, while Ep. 



124 HISTORY OP ENGLISH SOUNDff. 

generally has a. YP fluctuates between eoc^ eog and ee^ eg : it 
always has Heogun^ and generally spreocaUy while toegas is much 
commoner than wepas. North, agrees with WS: sprecan, toegas. 

441. This ^-smoothing is by the Germans called * palatal- 
umlaut'. The eNorth. arigfaraz^WS ear Af are, might, indeed, 
be adduced in support of a front pronunciation, but the 
bearvg=bearg of Ep. (438) points as conclusively the other 
way, and it seems most natural to suppose that *baarg be- 
came htBrg by absorption of the back vowel a into the back- 
modified r, and so with the other words. The spellings 
hnerh in Cp, and acerf in ^Y^hweoffy aceorf seem to point to 
a similar influence of/, which, of course, could only have 
rounded or backed, not fronted, a preceding cons. 

442. The origin of ea and eo^ as also of ea and eo, shows 
clearly that all the OE diphthongs must originally have had 
the stress on the first element, and there seems every reason to 
believe that in most of the dialects they kept it there through- 
out the OE period. There is, however, unmistakeable ME 
evidence of a shifting of stress (together, in the case of ea and 
<?(?, with a shifting of quantity) in unstrest syllables ; Orm's 
^ho o}*=0E Iieo, for instance, can only be explained from an 
OE hed through *hjd. This law of stress-shift in weak diph- 
thongs explains the INorth. am^eomi weak eom became first 
earn (435), then edm, and finally, by dropping the almost 
inaudible e, am. So also the earf of VP=WS eart appears 
in Du. as arj). In 1 WS we find the eo of teoro * tar ' as the 
second element of compounds passing through ea into a: 
ijigtearo^ ^taran. The same tendency to throw forward the 
stress in weak syllables is shown in the Olcel. shortening oiero 
* they are ' into re?, which, of course, presupposes ^erS, Jteir Sro 
becoming /^2> {e)r6 in order to avoid the immediate succession 
of two stresses. 

448. In MEt there is clear evidence of stress-shifting in 
all the diphthongs, whether strong or weak. That this 
shifting had taken place already in the OE period is proved 
by the alliteration in the poetical Genesis (text B) and 
Exodus, which are certainly of Southern, and probably of 
specially Et origin. In these poems we find such allitera- 



OLD-ENGLISH : VOWELS. J 2 5 

tions as junge, geome on ealle^ pointing to a pronunciation 
jomeyjalle. 

44A. In OE rinnan and hrinnan shift the ^ (510) and appear 
in eWS as teman, biemaUy IWS ymany byman^ IKt imariy biman^ 
but in Anglian (YP, Du., Ru.) they appear as toman {eoman^ 
eaman\ bioman {eo, ea) by the analogy of the older georn etc. 
The eo in eom (for *m) is no doubt due to the influence of the 
pi eorun. So also the late dat. pi Aeom for Aim is due to the 
influence of the gen. pi Aeora, the change being prompted by 
the desire to distinguish the pi Aim from the sg Aim. 

ft 

446. Answers to Gmc ai: Aid/, stdn, wdt * knows ' = Goth. 
Alaify stain, vaiU The second element of the ai was evidently 
weakened to e, and then absorbed. The analogy of the de- 
. velopment of Gmc au in OE (459) would lead us to expect 
^ai as the OE equivalent of Gmc ai, and this diphthong 
seems to be preserved in toeilawei (Cotton ms of Boeth.) fqr 
the ordinary tcdldicd * alas ' (Goth, vai * woe '), elsewhere icegla 
* euge ' with eg=ei (553). 

446. d before w answers to Gmc a (which otherwise re- 
mains in OE): sdwon *they saw', idwian * prepare ' = Goth. 
seAioun, tewa sb. Often also when a back vowel follows 
(cp d€eg, daga% etc): Idgon (and Uegon) *they lay ' = Goth. 
legun^ Idcnian * cure ' = Gmc *l€thidn (cp OE l^Bce * physician' 
= GotlL leieis), sldpan (and slaj[ian)^Ooih. slepan. 

447, In such words as swd, pd the d answers to Gmc a 
(Goth. wcL) in accordance with the general law by which all 
final strest vowels are lengthened (383). 



448. WS a corresponds regularly to Gmc m, which is thus 
preserved unchanged : far * danger ', of en, rad * advice ', sl^jpan 
= OSaxon/ar, dband, rdd, sldpan, Goth, slepan, 

448. In Kt and Anglian Gmc a is represented by e, 
thus Cp has fer, VP has efeUy slepan. There are, however, a 
few examples of ^ in the oldest texts: Ep. has ^/^=Goth. 
sve^ Aioar, naJfl=^WS nadl, blaeed=WS bl^d * blast', in all of 



126 HISTORY OF ENGLISH SOUNDS, 

which cases Cp has e, although in a few other instances it 
also has a, eNorth. has sua, ^^=WS ffia (461), Jksr. The 
eKt charters have stoa, Jtar, wcerony ^ad (in proper names). 
It seems as if the original a were kept when final and in 
proximity with r. VP has e everywhere. Ru. also generally 
has e, but a is not unfrequent, especially before r and after 
v> : mca, fiar, hwisr, radan are the regular forms, while in the 
verbal forms were, treron etc e is more frequent than a. Du. 
has 9U(B and occasionally tneron etc, elsewhere e, VP has 
only 9we,J)er etc. 

460. The Anglian e is constant in WS in mece * sword', 
which, as it occurs only in poetry, may be borrowed from 
Anglian. The -red of JElfred is sometimes written -ryd, point- 
ing to *-n^=:OHQ -rit (frequent in proper names), which has 
nothing to do with r^^=Goth. red. 

For the mutation-^ see § 481. 

451. In Southern ME shortened a is treated like OE a, 
showing that it must (in WS at least) have had the same 
sound, only long — x*. Lengthened ee in WS, as in %ade 
from scegde (401), must, of course, have had the same sound. 
Lengthened ^=x* ^^ C^* ^^ ^^ other hand, as in lede from l^gde, 
is never written (b. 



452. Answers to Gmc l\ win, stlgan, gellc. It is also the 
result of dropping Gmc mov n before a hiss or buzz (53i)» s^ 
mfif, sip *joumey'=Goth. Jimf, ainp. This * was, of course, 
nasalized at first. 



e 



453. Was originally a somewhat rare sound in OE. In the 
non-WS dialects it appears mainly as the representative of 
Gmc a (449), in WS (already in eWS) as an unrounding of 

a (489). 

454. The common OE e answers to Gmc S in her. In fnSd 

from Gmc *mezdd (Goth, mizdo), and also in the originally 
reduplicated preterites such as «&/7=Goth. saislep, it is prob- 
ably the result of contraction* 



OLD-ENGLISH: VOWELS. 12/ 

U 

455. Answers to Gmc u : Am, AlHd, ut. 

456. It is also the result of dropping a Gmc nasal before 
a hiss, as in cuj), ««=Goth. kunp, uns, 

457. In such monosyllables as Jiu, nil it is the result of 
lengthening Gmc u. 

6 

458. Answers to Gmc o, as in do, dom, moder. Also to Gmc 
S before nasals: mdna, mdnaji, cwomon 'they came ' = Goth. 
mena, menoj), qemun, and to Gmc nasalized a : pdhte^ han, fon (cp 
the ptcc hangen, fangen). 5 is further the result of the drop- 
ping of Gmc n or m before a hiss, as in djfery ff5s=Qoih. anpar^ 
gans, where we must assume the stages p, j«5, }*, }«. The a of 
Gmc ^tnano passed in OE through the stages X^^P^ pS running 
together with the vowel of QmcJ>dAta in its second stage. 

ea 

458. Answers to Gmc au, as in deap, Aiafod=zQoih, daupu9^ 
AatibiJ). In the oldest texts it is occasionally written ao, eo, aa, 
as in getksoi (Cp), eomc (Ep.), .Xodhald, JEanjied (BH), showing 
that the a of Gmc an became ^, in accordance with the general 
tendency of the language, the second element being opened, 
and finally unrounded. It is probable that the first element 
remained tB throughout the OE period; eEt, which often 
writes ia for eo, never uses it to express ea, 

460. Sometimes ea is the result of contraction, as in ^ from 
*aAwd (Goth. aAva) * water' through *ai(K)umy *auum, 9lean from 
*slaAan (Goth. slaAan). 

461. ea in WS also results from the combination c and g+ 
Gmc a in the same way as gaf become geqf (430) : sceap, giqfon 
'they gave', ^«rr=OSaxon scop, Goth, gebun, jer. The other 
dialects keep their tf=Gmc a (449): seep, gefon, get. gear^ 
however, occurs once in eNorth. 

462. In Anglian 9a is smoothed before c, g and A in the same 
way as ea (438). In the earlier texts the resulting vowel is 
written a. Thus Cp has (ml<Bc, bceg, AaA, with occasional geac 
etc. So also in eNorth., and in the earlier Kt and Merc 
charters. VP and INorth. have e\ ec, belec, ege, beg, ^i, neA, 



128 HISTORY OF ENGLISH SOUNDS. 



SO 



463. Answers to Gmc eu, as in ceosan, leof^ seoc^ deop = Goth. 
kiusan etc. Sometimes it is the result of contraction: seon 
from *seAwan (Goth. saiAvan), Jieon 'flourish' from ^Jjihan 
(Jneiidi in Cp). For eo in geomor see § 543. 

464. The original eu appears occasionally in the oldest 
texts : treutem, steupfadar (Ep.), utreum^ greut (eNorth.), 

Other early spellings are i«, as in gliuy sniuuip (Ep.), flitisum 
(eNorth.), ia, as in biafi pi (Cp, eNorth.), tiada prt 
(eNorth.). This last spelling occurs occasionally in VP, as in 
biap pi, gesiafi pi, and is common in eKt: bian vb, friand, 
bebiade, ie is occasional in VP : gesie vb, Jiend, Jiied^, The 
spelling io is frequent from the earliest period downward. 
Thus Ep. has biouuyrty kriosij), criopung by the side of streo^ 
hweoly beost, io also occurs occasionally in VP and in eWS, 
but afterwards eo becomes general in WS. ea occurs occa- 
sionally in the oldest texts, as in trea^ weadhdc (Ep.), also in 
VP, as in lea, gesean^ fead, and is frequent in INorth. 

465. In Anglian eo is smoothed into e before c, g and h. 
Thus Cp has thegh = ^h (WS Jjeoh), and VP has jlegan, geseh 
imper. This e sometimes becomes % as in tuigendi (Cp) = WS 
tweogende * doubting ', gefrigan * free \fligan (VP). 

466. That the difference between ea, eo and ea, eo was one 
of quantity is proved beyond doubt by the accents, the metre, 
and the whole history of the language. It is certain that 
the stress was not originally on the second element, for Gmc 
au and eu were certainly accented du, Su, The length must 
have been either on the first element, or else distributed over 
both. The former seems most probable. The lengthening 
probably began by an exaggeration of the glide between the 
two elements. Similar lengthenings occur in the OHG of 
Notker's texts, which write ie, ia, io^ 4o against iu, 60, 6u, eu, 
the circumflex indicating a long accented vowel, the acute 
a short accented vowel. 

Mutations. 

467. The OE mutations are all caused by Gmc i otj\ which 



OLD-ENGLISH : VOWELS. I 29 

even in parent Gmc had probably begun to modify a pre- 
ceding consonant (14^), the influence of the resulting fronted 
conss. on the preceding vowels being, however, carried out 
independently in the separate languages. In OE the Gmc i, J 
has often been lost, as in b^ndy s^ndan^Qoth, bandi^ sandjan. It 
must be borne in mind that -e and -ian in OE cause mutation 
only when they correspond to old -i, -^'an, as in bfnd, n^rian = 
Goth, nayan, not when they are modifications of a etc, as in WS 
girfen ptc= Anglian gefen (Goth. giban\ sealJian^OILQ salbon, 

9 
468. is the mutation of Gmc a, as in Aerede 'praised,' tellan, 
sendan^ weccan^ settan = Goth. Aazida, taljan^ sandjan^ vaijan, 
9aijan, As this mutation probably passed through the stage 
of X before it settled down to \ (or |J ?), there was a tendency 
to confuse § with a. The following words, for instance, have 
CB instead of § : baman, Aoffest^ laccan^ gerruBcca = Goth, brann- 
jan etc. f itself is often written a in the early texts. 

ie 

468. appears as the mutation of OE ea and eo (through 
i§^ and ty, i(B ?), that is, of Gmc a and e before certain con- 
sonant-combinations (4:27) : ierfe * inheritance ' = Goth, arbi^ 
ieldra^ compar. of eald, nieAt, &om *neaAti, Goth. nahHi)\ 
mer^ 'worthy,' hierde '(shepjherd* (cp weor} 'worth,* heord 
'herd'). K these words were formed by direct mutation 
from the Gmc forms, they would appear as evfe^ eldra; wirpe^ 
Airdeyihe mutation in the two latter being indeed already Gmc 
(299). The first two are, in fact, the forms that appear in 
all the non-WS dialects, except that VP has a {=a) instead 
of e before /-combinations : taalle ' well,' maltany aldra. VP 
has the Gmc >in afirran 'remove' {feorr 'far'), birAtu, Airtan 
' cherish ' {Aeorte ' heart '), but in other words it, in common 
with the other non-WS dialects shows the unmutated eo : eorre 
• angry * = WS ierrey Aeorde=WS Aierde. The eo in these cases 
seems to be due to the analogy of such words as the adv eorre 
and the subst. Aeord, where the eo is regular. Aeorde occurs also 
in Cp together with orfeormnisse (cp feorm) ; lurminburg, Aiordi 
in eNorth. eNorth. also has wiur^it from weorpin. In most 

K 



130 HISTOEY OP ENGLISH SOUNDS. 

caBes Cp agrees with Ep. in showing Gmc i, as in gesuirbet^ 
lirccy firgen. Ep. itself has the WS ie in geortoierdid^ orfiertme. 

470. There is another WS ie, which results, not from 
mutation, but from the combination ce-y ge- (534, 54^)) ^ ^ 
scieldy giefan, where VP and the other non-WS texts have 
the original e : sceld, gefan. 

471. WS also shows ie for f after c and ^, as in ciele * chill,* 
icieppan, giest=non-WS c^le, sc^ppan, gfst. The analogy of 
etese (483) shows that the ie in these words is not due to the 
direct action of the front cons, on the vowel, as in the case 
of scield etc, but that the ie is a mutation of prehistoric 
ea, itself the result of dipthonging a after a front cons. (535)> 
so that ciele has developed out of ^ceali, *cali, *iali, *kali (cp 
the strong vb calan). Hence there is no development of ie 
before § followed by a nasal, as in cennan^ cempa from *kannjan^ 
^kgnnjan ^kamjojo^ *ipmpjoj because Gmc a + nasal was not 
fronted in AFrisian and OE (415, 419), and so did not front 
a preceding back cons. 

472. WS also has ^ instead of ie before // from Gmc ^*, 
as in A^ll = Goth. Aalja, t^llan = taljan. WS has ie regularly 
before Gmc llj, as in fielhn from falljan^ tcielle irom *wallja. 
IWS syllan = Goth, saljan points to the WGmc doubling 
(3^5) which we would expect in halja etc; eWS, however, 
agrees with Et and Anglian in the form eellan. 

473. In WS front h before 8 and t mutates eo into ie^ as in 
siex * six ' = seox, Angl. eex^ cnieht = cnioAt, Angl. cneM. The 
eo of feohtan was probably preserved by the analogy of the 
other verbs in eo {weor^an etc) of the same conjugation. 

474. ie was no doubt a diphthong at first, but in the extant 
WS texts it must have become a monophthong, for it is often 
written simply t, sometimes y, which becomes general in IWS : 
ieffe^ iffey yrfe. The spelling erfe is rare in WS mss, and may 
be due to non-WS scribes. The change from the spelling ie to 
y is direct; without any intermediate i-period. The evidence 
of ME shows that this y had the same sound as the y in 
i^nn=f. The most probable explanation seems to be that ie 
was first smoothed to wide f , which was then rounded to f , 
in order to make it more distinct from the older % = I. 



OLD-ENGLISH: VOWELS. I3I 

y 

476. is the mutation of Omo u: fyllan^ 9ynn = QiO^,fulljan^ 
OSaxon %undiay gylden = OSaxon guldm. 
For 1 WS y = ie see 474. 

476. In IWS y becomes i before (front-modified) c, g, A, as 
in cicene, hricgy fiikt = earlier cycene * kitchen/ hrycg^ flyht. In 
some late mss (not in iSfcH) ci- for cy- is common, and cining 
occurs in Du., ki(ni)ng in Ru. together with the regular cyning. 

477. In the western IWS i in weak syllables becomes y, 
probably through X, as in ya, Ayne, Ayt, These forms are 
confirmed by ME West-Midland texts, which show such 
spellings as Aus ^Ynal fuse 'these/ where u^i (595). 

478. In IKt y becomes tf, as in aenn, gelden. The same 
change occurs in the IWS (MfcH) embe for ytiibe, and unnetlic 
= unnyttlic occurs once in eWS together with a few instances 
of embe. The change seems, therefore, to have begun in 
weak syllables, whence in IKt it spread to strong words, f 
was no doubt lowered to {, which underwent the same un- 
rounding as cs (479). Hence y in lEt is occasionally written 
for tf , as in cyrran = cfrran^ WS cierran. The change of y into 
e is shown also in the Suffolk charters of the end of the loth 
cent. : it was probably a general South-eastern change. 

470. is the mutation of 0, as in mle ' oil ' = Qmc ^olja from 
Lat. oleuniy dcsAter dat. of doAtor, As became u before t, j 
in Qmc (300), the regular mutation of Gmc is y, as in gylden 
(Gmc *guipina ; ^golpa = OE gold) ; a is the result of the 
mutation of a foreign (as in ode) or of the substitution of 
for u before the mutation began by the analogy of some 
other form, such as (in the case of dcsAter) the nom. 

480. fl? = { was unrounded to ^ [ not only in eWS (which 
preserves only isolated instances of os) but also in VP, which 
has ele, bledsian (where the cs was shortened from ce) through- 
out, the long ct being always kept in VP. 

481. is the mutation of a = Gmc ai in all the dialects : 
Aalan, anig (Adl, an =■ Goth. Aail^ ain), 

K 2 



132 HISTOET OP ENGLISH SOUNDS. 

482. In WS it is also the mutation of Qmc is : Uece^ dad = 
Goth, lekeis^ ded(i). In the non-WS dialects this a follows 
its original, becoming e: lice, did, 

le 

488. is the eWS mutation of ea (Qmc au) and io (Gmo eu) : 
Aieran = Goth, hauyan, geKefan = Goth. gcUaubjan (cp geleafa 
sb) ; cieH * chooses ' (inf. ceosan), ansien, gesiene = Goth. ananun(i) 
'visible.' ciese from Gmc ^katia (OHG k&H) points to an 
intermediate OE ^ceasi (535)* In the other dialects this ie 
appears as ^ .• herauy gesene, cise etc. Ep., however, has unhieri 
once, Cp alie^et once, and onsien is the regular spelling in 
VP, the other words having only i in VP. In many words 
io is retained unmutated in VP and the other non-WS 
dialects, as in stioran, getriowe=WS stieraiiy getnetce (cp 469). 

484. le is also written i in eWS, and rarely e. In IWS it 
becomes y : hyran, onsyny cyse. 

485. That le must once have been a diphthong in WS is clear 
from the originally dissyllabic hie * they,* ne subj., IWS Ay, sy, 

f 

486. is the mutation of «; bryd = Goth. irup{i), cypan 'pro- 
claim ' = Gk)th. Jkunfy'an through *cufy'an, ontynan ' open ' {tun 
* enclosure '). 

For IWS y^te see § 484. Before and, in some texts, after 
c and g it, as also original y^ becomes I (cp 476) in IWS: 
lean, blgan\ (n^in=eWS lecan, Inegan; cyj)an, 

487. In IKt y is lowered and unrounded to ^.' antenan etc. 
Hence y is sometimes written for i, a, as in lyce, lys9a = leee 
(WS l^ece), l€e8sa, 

488. is the mutation of : doman = Goth, domjan, soscan 
= Goth, sdkjan, gm (from *gami) pi ot gos. 

480. In IWS and lEt a is unrounded into i, a change 
which is already carried out almost completely in eWS as 
well : diman, sican, gis. 

Weak Vowels. 
400. In OE unstrest vowels were regularly shortened, as 



OLD-ENGLISH: VOWELS. 1 33 

in iunffena=Qo\ii. tuggono, Unstrest i and a^ which still occur 
in Ep., were afterwards levelled under e (=[?), as in pide^ 
tunge, Unstrest u and often interchange, as in m^igu, mpiigo, 

401. Such spellings as sagdig for 8€Bgde ic in Du. show that 
final vowels were dropt before another vowel in connected 
speech, at any rate in closely connected groups. 

402. Prehistoric OE i and u (answering both to Omc i, u 
and Gmc i, 0) were generally kept (in the later language as e^ 
u, 0) after short root-syllables, as in wine, 9unu, dropt after long 
root-syllables, as in wyrm, fdt-= Goth. vaurm(i)yfdtu. So also bap 
has pi bafiu, while Aus is invariable. If the long syllable before 
an « is half-strest or weak, the u is often kept, as in Jlscflddu 
compared with the simple flod, fern, and neut. mpmUcu ' human.' 

408. Dropping of medial vowels is frequent, depending 
partly on the character of the adjacent conss., and especially 
on the quantity of the preceding syllable, every unstrest 
vowel in a medial syllable followed by a single cons, being 
dropt after a long root-syllable, as in modrum, pigles compared 
wiih faderum, 9ta})olas. The dropt vowels are often restored 
in IWS, as in dferes = eWS o^es. 

The development of parasite-diphthongs has already been 
treated of (433). 

404. One result of the general dropping of final Gmc a, 
and the frequent dropping of Gmc i and u (492) in OE as in 
the other Gmc languages was that many words ended in 
syllabic vowellikes preceded by another cons., as in Goth, and 
OIceL ahtyfugly Goth, taikn from *aira, ^fugla, ^taikna. In OE 
syllabic n and m, and / after forward conss. are generally kept 
unchanged, as in tdc(e)n^ toaHm, nadl, husl, while syllabic I 
after other conss. and syllabic r after all conss. develop 
a parasite vowel — u (later 0) after a back, e after a front 
vowel in the root-syllable: fugol, dtur (dUor); atppel, winter 
= Goth, vintru. Ep. still preserves atr = Olcel. eitr etc. 

405* The insertion of an f between ^ = d> and a preceding 
cons, is regular in byrig for the rarer byrg, and occasional in 
other words, such as fylgian ' follow.' u, is sometimes in- 
serted between r, / and a following cons., as in burug for bmg, 
kehtstr (433). 



134 HISTOET OP ENGLISH SOUNDS. 







Consonants. 








406. The 


following 


was the normal OE consonant-system: 


TflB5AT. 


BACK. 


FBONT. 


lOBKW. 


LIP. 




h 


h 


h 


hr 


>.8 


f,hw 






— 


— 


hi 




— 






c 


• 

c 


t 


— 


p 






— 


— 


hn 


— 








g 


m 

g 


r 


h^ 


f,w 






— 


(0 


1 


— 


— 






e 


g 


d 


— 


b 






n(g) 


(n) 


n 


— 


m 




s 


c 


o 


o 

CO 

o 


w,s 






a 


Q 


— 


D 






— 


— 


1 


— 


— 






e 


<D 


(D 


W, S 


>, o 






— 


(CD) 


CO 


- 


- 






6 


ffl 


XD 


— 


D 






J 


w 


1 


— 


r 




Observe the ambigu 


ousness 


of 












h=». 


c, o. 












c=a, 


Q. 












g=e, 


<D, Q, ffl. 












n=d, 


7. 










h 


8, f=u, 


S, > ; W, 8, >. 






Note also that r, i 


',».«] 


jrobably 


had a 


front or 


front- 


modified sound before 


dand^ 


• 









h. 

407. In OE, as in the other Omc languages, A was 
weakened to a mere breath initially. This is proved by 
the occasional omission or addition of an initial k, which 



OLD-ENGLISH: CONSONANTS. 1 35 

occurs throughout the OE period. Already in Ep. we find 
HBnl as well as AaM, Aynnilac s= the correct ynnelac of Cp. 
So also in gihiodim, = geeodan, where it is practically initial 

498. Medial A before a vowel (especially between vowels) 
was not only weakened to a breath, but completely dropped, 
the resulting hiatus being generally got rid of by contraction. 
Ep. still preserves the h in such forms as smhorM = WS 
sweoTM 'fathers-in-law/ fwrhum = Vf^ furum^ dat. pi oi fv/rh 
* furrow.' 

400. Medial h is also dropped before the vowellikes r, /, », 
m, «7, as in neaUkcan ^ approach ' = ^neaAlacan, dwer = d-itcar^ 
which latter form is, indeed, often restored (as in other words 
as well) by the analogy of the uncompounded Aicar etc. 

500. The dropping of h in ahw&r etc is really part of a 
more general law by which the breath-^ was regularly dropt 
in unstrest syllables. This is clearly shown in proper names, 
such spellings as JElfere^ Eadelm for JElfherey Eadielm being not 
unfrequent even in early texts. The history of the pronoun 
it in ME (724) makes it tolerably certain that OE must have 
had the same distinction as MnE between strong Ait^ him^ heora 
and unemphatic "^it^ *%m^ eora. eara, indeed, occurs several 
times in Bu. 

601. Initial h before a (vowellike) cons, in the combinations 
Ar, hly Ato, An, as in Aring, Alaf, AvxBt, Anviu, must in Gmc have had 
its regular sound c, for not only does the A of AUf etc (Ooth. 
Alaif) answer to an Ar. h (Lith. hlepa%\ but the Qmc *Alaiva 
itself was adopted by OBulg. in the form of cAlebU ccD[fDl. 
The next stage was the reduction of the ^ to a breath. This 
stage, in which A and / etc were pronounced separately, is 
preserved in the laws of alliteration, by which Al alliterates 
on the A of Adm etc, and also in the shifting Aon from ^AroM 
(510). In MnE these combinations were merged in simple 
r etc, except that wA still partially survives as a voiceless to. 
It is quite possible that the OE Ar, Al, A10, An were really 
simple o, C9, o, 1, for the alliterative usage may well be only 
traditionaL This is supported by the spelling rAing in Ep., 
although the rA may be due simply to the analogy of Latin 
spellings such as rAetor. 



136 HISTORY OF ENGLISH SOUNDS. 

602. h kept its Qmc sound when final, when doubled, axid 
before conss. in the combinations x (=A«), ht^ as xnfeoh (gen. 
feo9 from '^feohe9\ fearh (gen. feores), furh (dat. pi furum) ; 
kliehhan; oxa=-Qoih., auAia^ beorAL In the oldest texts A=c 
is generally written cA (except in x) : thus Ep. has tocA, fiorcA^ 
tore At =WS toA^purA^ tor At * bright.' AA is also represented by 
cAA^ AcA. cA is sometimes abbreviated to c: thus we find 
dcfripu in a Runic inscription, and ^a/c=WS AealA (sinus, 
' -hale ') in BH twice. In the combination ^ct the dropping 
of the A is more frequent than its retention : thus Ep. writes 
necti^ala, torctendi. One Runic text has unneg^ fefftqf=WS 
unneaAy feohtap parallel to bt^ft (515). Cp has slag^ slagA^ 
mispagcA^sldA, misJ^A, These last spellings are compromises 
between A and ^=€. 

503. These usages are no doubt mainly the result of Celtic 
influence. In Olrish and OWelsh A had no independent 
value, being mainly used to fill a hiatus or mark emphasis, 
and in OWelsh c was often written instead of cA. Hence the 
eOE prefixing of A, its use as a hiatus-filler (as in giAiodun\ 
the hesitation to employ A to represent c, and the shortening 
of cA to c. The later use of A everywhere may be due to 
the same Runic influence which superseded tA by /, and uu 

504. In WS and Et, when A comes before a hiss or buzz 
(/ or *) by the dropping of a vowel, it is preserved in the 
form of c or o, as in WS {ge)9ieAst^ sieAf from ^riAtois^ ^siAvnpy 
niehut^ but dropped in the Anglian dialects : thus VP has gesis^ 
gesipy neat, 

505. In At the A must have had the front sound o in WS, 
for it mutates a preceding ea^ ^0 to i^, as in niekt^ cnieAt^neaht^ 
cneoAt, which occur occasionally in eWS. 

r,l. 

506. The OE r was no doubt a strong point trill as in the 
present Scotch dialect. 

507. The parasiting influence of r and ^(427) shows that they 
were probably formed as in MnE — with concavity of the fore 
part of the tongue — which gives them a kind of guttural 



OLD-ENGLISH: CONSONANTS. 1 37 

quality favorable to the development of a back parasite- 
vowel, which, if uttered muffled — ^with imperfect lip-open- 
ing — is easily rounded, r and / cannot have been full back, 
or even back-modified, conss. — €i, 8 or (D(, qo(— because in 
that case single r and I would have diphthonged a preceding 
vowel, instead of requiring to be doubled, or to have the 
support of a following cons., the effect of which probably was 
to lengthen the r or / and so increase its volume of sound. 

608. Before c and ^, r and I probably had a front-modified 
sound, as in wyrcan^ swdc, 

609. r and / answer to the Gmc r and / resp. But r is 
also the representative of Qmc £^, as in gecoren from Qmc 
^iozandy ifrian^Qotix. hazjan 'praise,' mierran * hinder '= Goth. 
marzjan^ hord^Qoih. huzd (145, 315). Final Gmc z is dropped 
in OE ; hwd^ md * more * adv from Qmc *Awaz^ *maiz. 

610. r is often shifted from before to after a vowel, when 
this vowel was followed by nn or «+cons. (that is, by breath 
«), as in bumay Aors, berslan=^Qoiii. brunna, OSaxon kro99^ OHG 
breitan. The original double conss. are still preserved in the 
spellings bumna, horssum (0£T). There is shifting before 
simple 9 in g^trs. That these shiftings are comparatively late 
is shown by the frequent occurrence of the unshifted forms 
in the oldest texts: gnBs^ rendegn^WS gar9, <Bmpegn. The 
shiftings birdas,j^rda=zWS bridd{a9), Jridda are INorth. 

611. There are some shiftings which, in the earlier period 
at least, occur only in unstrest syllables, especially the second 
half of proper names. The earliest is the change of -frif into 
-frf, "ferpi Tidfirp occurring in an inscription. Then follows 
the reverse shifting by which 'be(o)rM becomes -brekty CeolbreAt 
appearing in a WS charter of 778. Afterwards -breAl passes 
through •*brieht (505) into bryU in WS. In the INorth. brekt 
by the side of berkt we see the shift carried out in the isolated, 
strest word as well. INorth. has a similar shifting mfrokiia^ 
fryktu by the side of the olAfft forUia^fyrktu. 

612. Shifting of / occurs in unstrest syllables. Regularly in 
the ending -id^ still preserved in the gyrdul of Ep., lAtergyrdeh^ 
and in -giil (= strong gisl ^ hostage') as the second element of 
proper names, as in Cynegih= Cynegisl, which also occurs. 



138 HISTOBT OF ENGLISH SOUNDS. 

518. The dropping of r in specan, 9pac:=^sprecan, sprSec is IWS 
and lEt. 

h B, f. 

- 614. These conss. were probably formed exactly as in MnE. 
It is, however, quite uncertain when / changed from Gmc o 
to the present >. 

515. OE f and s always correspond to Qmc / and «, for 
Gmc 9 and z became d and r in OE (315). OE^, on the 
other hand, corresponds both to Gmc / and Gmc 1?, which 
latter first became WGmc d, and then 9 (>) again in OE. To 
the first/ corresponds an OHG/, t;, to the second an OHG b 
(326): wuIf^OKQ wolf (Gmc ^w6lfa)y hafen ptc=OHG haban 
(Gmc *havand). In the oldest texts the symbol/ is restricted 
to the former /; gimfa (OHG grave) ivulf the latter being 
denoted by d, as in scribvn, salb (OHG 9cribun^ salba). It is 
clear that this b denotes an open cons., for u is written in 
nuida (Ep,)=:9ifipan 'siftings' in Cp, and in Cp we find 
ll=ft in lybL The use of medial b to denote a r-sound is 
probably due to the popular Latin of Britain, in which the 
medial b of such a word as habere had this open sound. We 
have direct proof of such an open pronunciation of Latin d 
in the spellings Leonipi^ Marpmius etc, which are the regular 
ones in Or. The later general use of/ is due to the influence 
of the Bunic alphabet ; perhaps also to the Celtic use of /=>, 
as in OLrish and in Mn Welsh. 

516. fi is generally denoted by the Latin th in the oldest 
mss, especially initially and finally, which th is sometimes 
abbreviated to t (cp 502): thus Ep. has ihegn, loihiy latk^ 
earbetiic (=WS earfoplic). Medial f is often denoted by rf, 
which also occurs finally, and even initially: thus Ep. 
writes 9ceadan, giroedro^ uueard^ gidopta, dislum=:WS sceapan^ 
gerepray wearf^ gepofta^ pklum. The Bunic } is rare in Ep. : 
}m9^ suipiBy milcip, as also 9i ffinga^ qutSa^ mvf, Ep. has once 
dh in fordh. The blendings fik {suaphe OET) and 9h (9Auehl 
Cp) also occur. In the later Anglian texts 9 is universal, 
as also in most of the oldest WS and Et mss. In some eWS 
mss^, however, predominates. In IWS there is a tendency 



OLD-ENGLISH I CONSONANTS . 1 3 9 

to write >& at the beginning of a word or letter-group, 9 else- 
whera The distinction was no doubt purely graphic, p 
having the character of a capital and therefore being a good 
initial. 

617. Gmc ip becomes Id in OE, the older forms being still 
preserved in the earliest texts : thus Ep. has haWi^ 9cyi%haU^ 
(t=tA) as well as toAald, Cp h&a feUAay early Merc, charters 
have -iall (where i=tA), 

618. That s between voiced sounds was voiced in prehistoric 
OE is proved by such contracted preterites as /i««^=Goth. 
lansida compared with cyifle=*(y9ste from cyasan^ for if the 9 of 
Uesan had been voiceless it would have changed the d of the 
ending into t in the same way as S9 does. 

619. 9 is often shifted in the medial and final groups 9c and 
4P, especially in IWS, the shifting having apparently begun 
medially : dc9ian^ dxiatiy cirp9 scolder d9cian^ ^crisp. 

620. We see that there is decisive evidence that intervocalic 
}> and 9 were voiced in OE. In Gm. and Dutch we have clear 
evidence that initial fi was also voiced in such a word as ding 
^OEtJ^ing. In Dutch initial 9 is also voiced, as in zeven=OE 
9ecfon, and this has been adopted by the High German of the 
North — the Upper Gm dialects still keeping voiceless * every- 
where — medially as well as initially. OHG had initial v, as 
in voli ; and this v is stiU preserved in Dutch, while it has been 
unvoiced in Gm, which, however, still keeps the old spelling. 
The evidence of ME and of the Mn£ dialects shows that in 
the 12th cent, initial 9, z, v must have been fully developed. 
It seems therefore plausible to assume that the Gmc ^ and 

/ were voiced initially as well as medially in WGmc and, 
that the initial voice of HGm ding and its predecessor 9ing 
(which must be at least as old as the 7th cent.), together with 
that of the corresponding Dutch and Southern E. forms, are 
not independent developments, but remains of a common 
stage, 9 following the analogy of the other hisses in Low Gm 
and Southern E. 

621. In Gothic, on the other hand, it is certain that the Jf 
and/ were voiceless in all positions — intervocalic as in qijkny 
as well as initial and final, as in /nufi — for intervocalic w and 9 



I40 HISTORY OP ENGLISH SOUNDS. 

were expressed by d and b, aa in naubaimbair =Li november. In 
the Scandinavian languages Jb and / are breathed initially, 
voiced medially and finally, as in Mnlcel. Jfin^, kveSa^ ver9; 
fara^ hafa, haf^ s remaining unvoiced everywhere. The pro- 
bability is, therefore, that initial hiss-voicing has never taken 
place in East-Qmc (except in Sw-Dan. det etc, where it is due 
to want of stress, as in E. the etc). It is further possible that 
the Anglian and Jutish (Kentish) dialects of OE, which were 
geogi*aphically closer to the Scandinavian languages than the 
Saxon group, may never have developed the initial hiss-voicing 
of the latter. The evidence of /and b in Ep. would, indeed, 
prove that even medial hiss-voicing, so far from being common 
WOmc, was of comparatively late date in OE. But the whole 
evidence beaiing on / is entirely the other way, showing that 
voiced p was fully established in Ep., not only medially, but 
probably also initially. This conflict of evidence makes it 
possible that the distinction between /and b was ideally meant 
to express that of > and 3. Cp also bb fromjg' (557)- 

622. Final/ and/ are always voiceless in Gm and Dutch, 
but this is merely the result of their general tendency to un- 
voice all final conss. except the vowellikes. In the Scandina- 
vian languages, on the other hand, they are always voiced. In 
the present Southern E. even the final s of such a word as OE 
gd9 is voiced, and there seems every reason to suppose that 
final /,/, * were voiced in OE also — at least in WS. 

628. In such combinations as st, ft the hisses were, of course, 
always voiceless. The latter combination is in Ep. often ex- 
pressed by pty as in saept^fs being also expressed hj ps and b^ 
(cp lybt 515) in arapsid, rabsid. This use of pt (which is also 
Olcel.) may be due to a pronunciation of Lt pt in capttc9 etc 
as/^. M and jf were always voiceless. Ep. has waffsa9=zWS 
waspas, foapsas, ff occurs only in the foreign offrian, Offa and 
some obscure words. 

624. There is an OE law by which in the combination 
voiced stop or buzz + buzz both elements are unvoiced. Thus 
bledtian in VP (from ^blodizon) becomes bkisianin WS. So also 
eWS guhunff becomes gltsung. In the North. LY this law is 
carried out regularly in compound names such as Altfrip 



OLD-ENGLISH: CONSONANTS. I4I 

(=^Aldfnfjy Eatfrip^ Eatpryf, This tendency is evidently the 
result of the attempt to strengthen the acoustic effect of the 
open cons. Of course, if either element is voiceless, the change 
is still easier, as in mitsceatt=MedsceaU. 

626. / in the combinations tp, (//, {^ is first unvoiced together 
with the preceding d and s, and then stopped, giving t(t), st, as 
in WS bitt^bilepy bidep, ciest=*cieifep, and in hafa9iu^hafa% fu. 
Hence in WS ip is often written instead of radical %t^ as in 
fa9p ^fasi. So also regularly in fatte ^fat fe^ and occasionally 
mfat UBty fat td etc, showing that in actual speech the initial 
/ of pronominal words was regularly assimilated to the ^ of a 
preceding /<9^, a change which is consistently denoted in the 
ME Ormulum. fs is smoothed to ss in later OE, as in blissian 
from older bRpnan. 

626. The change of final f into 8 in verb-inflections in 
INorth., as in bindes, bindas=VfS bint (Angl. bindef)^ bindap 
seems to be organic, as there do not seem to be any analogical 
influences at work. 

w 

627. is expressed in the oldest texts by uu, as in uueg, single 
u being generally written after a cons., as in cuic. In North, 
single u is preferred everywhere : uerc. But already in Ep. 
there are a few instances of the Runic Wy which became general 
in the course of the 9th century. 

628. The OE w must have had the same sound as in MnE, 
nl that of a consonantal u. 

629. Final w after conss. is vocalized to 1^, (?, as in nearo from 
Gmc ^nanca through *nartc. After short vowels final 10 is 
vocalized, and contracted into a diphthong with the preceding 
vowel, as in eneo * knee ' from Omc ^Imetoa, The w is often 
restored by the analogy of other forms with intervocalic w, as 
in the prt cneaw * knew ' by the analogy of inf. cndwan. So 
also eneo becomes eneota in WS by the influence of the pi 
cneawu. After long vowels and Qmc diphthongs final w is 
regularly dropt, as in «<?=Qoth. sdiw{i\ a = Goth, aiw^ but is, 
however, often restored by the influence of the inflected forms, 
as mindw. 



142 HISTOEY OF ENGLISH SOUNDS. 

H) ixl 

630. n had, of course, the back sound before c and ff, as 
in sincan, singan^ and the front before 6 and ^, as in sfncan^ 
ipigan, 

581. Gmc n and m were dropt in AFrisian before the hisses 
and buzzes/, «,/with nasalizing and lengthening of the pre- 
ceding vowel, a becoming o (458), as in mu^^gas^flfirom, *munp, 
*gans, *fmf. 

632. In North, final weak n was dropt in inflections ; already 
in eNorth. we find galgu, gistiga=zWS gealgan^ gestigan. The n 
of the past partic. (gestigen) was not dropt, because of the 
influence of the inflected forms geetigene etc. 

Stops. 

633. There is a tendency in OE to unvoice final voice stops 
preceded by a vowellike, when unstrest, as in eWS sint, which 
appears originally to have been the weak form corresponding 
to the strong nndon^ the ordinary sind being a compromise. 
Other examples are eWS weorfmi/nt, elpent^farelt:=^weor^yndy 
elpend^ftsreld, Cp has /uelsent *augurers/ So also final -ing^ 
~ung are sometimes written -inc, unc in early texts : wlatnnc 
(Cp), Cymesinc^ together with the compromises -ingCy -incg etc. 
These ^s and cs are often carried over to the inflected forms : 
fareltey gestincum * guestings,' * exiles * (Cp). Conversely, we 
may assume that the later uninflected forms ftBreld etc owe 
their ^s to the inflected forms. There is no reason for sup- 
posing that the final conss. of fully strest words were unvoiced : 
such spellings as felt in the older texts really stand for felth 
etc (516, 517). 

c. 

634. The history of the mutations shows that all the Gmc 
conss. were liable to fronting when followed by j or i. In 
most cases the fronting was afterwards lost : thus in such a 
word as ^nde it has left no traces besides the mutation of the 
preceding vowel. The back conss. c and ^, however, have pre- 



OLD-ENGLISH : CONSONANTS. 1 43 

served a modification of the original fronting to the present 
day in such words as cAin, nnge (sing) firom OE cinne, 9pigan^ 
contrasting with kin^ sing from OE cynn, singan, MnF. (J) and 
(ds) are, indeed, unfailing criteria of OE fronting. 

635. Initial Qmc k became c in OE before all vowels which 
were front before mutation set in, that is, before a^ «, e, ea, eo, 
^=Gmc a and its mutation, i, e, ia, eo, and remained un- 
changed before the back vowels a, u, 0, ^=:Qmc at, u, and the 
mutations ^, y, a?, a mut. of a=Gmc ai, y, cs. In WS ca, ce, 
ca become cie, cie^ cea resp., cea^ cea being liable to mutation 
into cicy die resp. Medial Qmc k became c before i andy, as in 
scsdan^Goih. sokjan. This c (which often becomes final in OE 
through dropping of a vowel, as in the imper. 9^6) is, of course, 
always preceded by a mutated vowel. In the early texts, 
especially in eWS, it is often denoted by a following i or ei 
bircia (Ep.), gescindo (Cp), leceaa (Ep.), recceo (Cp). eWS 
generally has e: secean (and secan), recc(e)an, but also iy 
especially before u : ecium. In INorth. and IMerc (Du. and 
Ru.) there is no trace of these es and /s, and they, are rare in 
IWS. But while lAngl. writes only ffeTica, pencan etc, there are 
instances of fencean etc even in quite late WS texts, such as 
the Gospels. Taken in connection with the ME evidence (741) 
the lAngl. spellings seem to point to a return to the back c in 
secan, pencan etc. The spellings ekan, besenked in Ru. seem, 
indeed, to be decisive on this point (538). The return to c 
may have begun before back vowels, as in the infin. sotcan. 
The absence of intrusive ^ , i in VP may be simply due to that 
striving for brevity which is characteristic of this text, being 
shown, for instance, in its regular omission of final cons, 
doubling (409) and its want of accents. 

686. There was a similar return to back e and g in Angl. 
before a/ + cons., as in cdld^ gaile =zWS ceald, gealle. 

637. The evidence of ME shows that c was fronted in the 
combination hc before all vowels, the foreward s evidently 
drawing the c forwards. The c of sc follows, of course, the 
same laws as simple c before AFrisian a etc, as in WS sceal== 
VF seel, and in these cases the e is always written. But see is 
also written before originally back vowels, although here it is 



144 HISTORY OP ENGLISH SOUNDS. 

often omitted, especially in the earliest texts. Thus WS and 
Du., Ru. write sceadan for the earlier giscdd (Ep.), scddan (VP), 
eWS and Ru. frequently omitting, IWS and JDu. generally 
keeping the e. In IWS sco, sco generally take the e : sceolde^ 
sceorty 9cedc, while scu, scu are generally written sceo, sceo^ less 
often sceu, sceu, and occasionally simple seu: seucca, seeticca, 
sceocca (MtcH\ scufan, iceufauy 9ceofan. The e was no doubt fully 
pronounced in 9ceal etc. Indeed, if this word had not been 
pronounced with a full diphthong, there would be no reason 
for its IWS form 9ceaU, evidently due to the analogy of eall, 
weall etc and the rareness of -eal. The frequent omission of 
the e in the other class of words {iceddan etc) seems to show 
that it was a mere diacritic in eWS. But the IWS change of 
eu into eo makes it probable that it had developed into 
the first element of a diphthong with the stress on the first 
element, which, in the case of ed^ took the length from the 
second element. There is ME evidence that sceadan from scddan 
had the same diphthong as bread etc. 

538. In the oldest texts, and occasionally also in Ru., cw is 
expressed by the Latin qu : quidu^ quoen. More important is 
the occasional use of ^ to denote the back c, as opposed to the 
front c, which, again, is exceptionally frequent in Ru., and is 
not uncommon both in early and late WS. Thus Cp has iali^^ 
iylle, LV has Kanla {=.CpUa\ Kcsna, Fronka^ Ru. has kdsere^ 
hyning^ king^ ciken ' chicken,' and kynn . kinn etc are frequent in 
WS. ck also occurs, as in Backa (LV ). The distinction may be 
due to the Runic alphabet, in which the cen-rune seems origin- 
ally to have been restricted to c, the >(-rune formed from]['gar 
(itself originally = back g) being used to denote the back c. In 
the actual inscriptions k is restricted to the back sound : kristy 
kyning^ bekun^ but the more frequent cen is occasionally used to 
denote the back as well as the front sound : becn^ Sac, cup. 

530. The analogy of g (553) and the ME evidence (742) 
make it tolerably certain that c was often fronted after a front 
vowel, as in ic, nopllce, 

540. Final c in unstrest syllables often becomes h in North. 
We find meh in a Runic inscription, meh^pehy mih in INorth., 
together with ih, which is also written^enclitically ig, as in 



OLD-ENGLISH : CONSONANTS. I45 

9agdig = fagde ic. This ig seems to be simply another way of 
expressing the same sound, for we find -iA written for un- 
stressed -ig in INorth. aA for ac occurs also in VP. 



g 

541. Gmc g (including earlier Qmc j) splits up into g and p 
in OE according to the same laws which govern the distribu- 
tion of c and c. Of the two Runic symbols gefu and gar, the 
former probably denoted g, the latter g, but they are not clearly 
distinguished in the existing inscriptions. 

542. Initial g became g=m before a, i, e, ea^ eo^ m = Gmc <i, 
f, e, ea, eo, and was kept unchanged before a (= ^), 1*, 0, a = Gmc 
ai, «, 0; f, 5^, fl?, ^ = mut. of «, y, a. In WS ga, ge, ga become 
gea, gie, gea resp., gea, gea being liable to mutation into gie, gle 
resp. 

543. Initial Gmc^ was hardened into the stop m in OE, and 
was thus confounded with g both in sound and spelling. It 
was expressed by simple g before front vowels, as in ^i/=Goth. 
jahai, Gessu^ (Bunic inscr.) = Jesus, Anglian ger = Goth. jer. 
Gmcy« becomes gea in WS, as in gear. Before back vowels it 
is often expressed by gi in the older texts, but generally by 
ge, as in gioc^ geoc = OHG joA, gedmor = OHG jdmar (458), 
Giupeas ^ Jews ' in a Runic inscr. = the usual half-Latinized 
Judeas, giung (Du. Rit.) = Goth, jvgg. In WS geu becomes 
geo (cp sceocca from sceucca 537), as in geong,geogof. VP gener- 
ally writes simple g in gung, gugup. The Latin i is not unfre- 
quent, especiaUy in Ru. : iung (also in VP), ioc (also in Kt ch). 
In eWS we find the forms imig, giung, giong, geong, the last 
becoming general in IWS. Such spellings as iung, iu in IWS 
seem due to Et or Mercian influence. The spellings gung 
and ivng show that in the nonWS dialects g from j had 
no more diphthonging influence than g from Gmc g, while 
the WS change of *geung into geong shows as clearly the 
development of a full diphthong with the stress on the first 
element. 

544. Even g = Gmc g is sometimes written L Thus Cp has 
ieces = geaces, and Ru. has regularly iarwan = geartriau and ierd 

L 



146 HISTORY OF ENGLISH SOUNDS. 

= gfrd^ WS gierd. Even in foreign words we find lorius = 
Oeorgiiis (OET). 

646. In the poetry the two ^s alliterate freely not only with 
one another but also with g^ thus geong alliterates with giedd^ 
Oeat and with gold, 

646. This last fact is generally cited as a proof of initial g 
and g being open conss. in OE, as in MnDutch, where geven and 
god are both pronounced with e or c. It is assumed that such a 
word as god in OE had initial e, while geofig, grfan had initial 
<D, so that instead of Omc^' being hardened into a stop, it was 
Gmc g before front vowels that was levelled under ^\ 

647. Plausible as this theory seems, there are fatal objec- 
tions to it. The WS change of ce into cie is the result of the 
almost inevitable development of an open front glide, which we 
may roughly call^', between the stopped front cons, and the 
vowel, and if we assume that in ge the g was also a stopped 
cons., the change into gie is perfectly analogous and intelli- 
gible, while that of *je into *jje is unmeaning. The same 
argument applies equally to g from Qmc^': iS giung meant 
simply J«»^, the development of a^'-glide would be as unin- 
telligible as that of a «^-glide in such a word as willa^ the open 
] and w being themselves practically glides. Again, LV 
writes Eadgar^ Aldgul etc, but if the g were really an open 
cons., we should expect to find the preceding d^ become t 
(524), which is not the case. Another argument in favour of 
the change of ^' into a stop is the loss of the Runic j and the 
use of gefii — which must certainly have originally denoted a 
stop — to represent both Gmc j and the OE fronted Gmc g. 
The use of t in iung to denote a stop is in complete harmony 
with the Late Lt pronunciation, in which, as the Romance 
languages show,^' must have become a stop (88). 

648. In ME all initial g^ became ^ (745). The alliteration 
proves that in OKt this pronunciation was already established, 
not only in jung but also in ^jorne = geome, etc (443), so that 
two such words as geald and eald were both pronounced jaldy 
whence the not unfrequent confusion between ea- and gea- in 
some late mss : eartoe for geartoe, gearfope for earfofie. This j^ 
sound was, of course, as naturally expressed by i as the older 



OLD-ENGLISH : CONSONANTS* 1 47 

g was, and it is probable that the iarwan of Bu. really meanB 
jancan, with absorption of the e. 

549. The same weakening seems to have been carried out 
much earlier and in all the dialects in unstrest syllables. 
The prefix ge- (older ffi-) and the pronoun ge never insert an i 
in WS, as they would if the cons, had been a stop ; ffie in Du. 
may be an emphatic form of ge. ge- is written ie- twice in 
old mss. i for g in the second element of eEt names such 
as ^fiiliterd = ^^Igeard may also represent weak J. 

650. Uninitial g was a stop in the combination ng^ as 
proved by the final change into nc (533). ng was, therefore, 
da, as in MnE longer^ after unmutated vowels, as in singan, lang. 
When preceded by a mutation — that is, when followed in Gmc 
by i OTJ — it had the sound lcd, as in s§ngan ' singe ' from Gmc 
*sangjan, Ipig. 

551. Uninitial g was also a stop when doubled = Gmc gj. 
This doubling is written gg in Ep. and occasionally in later 
texts, but the usual spelling is cg^ thus Ep. has eanoigga^ mygg 
= later eancicgay mycg. cgg^ gc^ gcg are occasional variants. 
Shortening of final eg into c or y is very rare. As this group 
must necessarily be always front (l^cgan = f^ggan, Goth. 
lagjan), it is probable that the c was introduced in order to 
indicate this front quality, there being no special letter for g. 
This is confirmed by the frequent use of gg in the few 
(probably foreign) words in which the doubling occurs after 
unmutated vowels, and therefore expresses gg, not gg, as in 

frogga * frog,' clugge * bell,* which last is certainly Celtic. It is 
also possible that the combination eg was meant to symbolize 
a half-voiced or whispered gg, for we find styphum for dyhbum 
once (GET). 

552. Elsewhere uninitial g was an open back or front cons, 
(e, (d), the open g occurring under the same circumstances as d 
(535)> *"^d like it, being expressed hjg,gi and ge. The open- 
ness of the g is shown by ME and the evidence of the OE 
sound-changes and spellings, which will be treated of further 
on, and is made a-priori probable by the fact of a similar 
change having taken place in all the MnGmc languages 
except Upper Gm (330). Open g occurred not only after 



148 HISTORY OF ENGLISH SOUNDS. 

vowels, but also after the vowellikes r, /, as in beorgaHy 
Jyigan. ^ 

663. Open g is, of course, always front before Qmc i, j^ as 
in Mfnigo = Qoth. managei^ fy^g^n, from ^fulgjan^ being often 
expressed by gi^ ge in the same texts which write ci^ ce for c : 
h^giung (Cp), wfuigeo, Jylgean. By a later change open g 
became front after front vowels when final or followed by 
another front vowel, as in dag, weg^ gen. dages, tceges^ grag. 
So also after the Anglian smoothings of the WS diphthongs 
as in ^^^ = WS eage. In Rt, indeed, this g is frequently 
written i. Already in Ep. we find grei etc, and dei is 
frequent in the later Kt. The North. LV has Meiuald 
once = WS Magweald, These spellings do not occur in WS, 
but in IWS such spellings as daig are not uncommon. 
They may be regarded either as showing the development of 
a glide — (DircD, or as compromises between the traditional dag 
and the phonetic ^dai. The latter view is the most probable : 
there is every reason to believe that the Kt spellings represent 
the general OE pronunciation, and that g preceded by a front 
vowel had sunk to a diphthongic vowel, or, at any rate, had 
lost all consonantal buzz. If it had preserved any consonantal 
quality, it would have followed the analogy of final open g 
(554), and become o, and would have been preserved as a hiss 
cons, in ME — neither of which is the case. The use of ^ to 
denote i (376) shows, too, that even in eWS medial and final 
g had been completely vowelized — after i at least. When final 
g had once been weakened into a vowel, its parallelism with c 
was lost; hence it did not revert to the back quality in 
Anglian, as was apparently the case with uninitial c ( 535). 
There is, however, ME evidence of such a reversion in the 
case of medial g (750). Open g preceded by a front vowel and 
followed by a final vowellike is always front, apparently even 
when a back vowel is added in inflection, as in 9egl^ regn^ 
generally written segel^ regen later. 

664. Open g is necessarily back finally or medially after a 
back vowel, as in irog^ geiwg^ boga, gebogen^ also when r, I come 
between, as in bnrg^ gealga* In IWS and IKt final open g is 
written ^, showing loss of voice : troh, gefioh, burh. The spell- 



OLD-ENGLISH : CONSONANTS. 1 49 

ings gh^ hg occur occasionally for final open ^, as in slogh^ (utahg 
{gh already in Cp), the latter even for medial open g^ especially 
after r, as in hurhga pi, beorhgan, hg is also used for g between 
front vowels, as in wihgena gen. pi, getoehgene^ though not for g 
preceded by a mutation. Some late mss occasionally write 
simple ^, as in eME. Open g seems also to have been back 
after a front vowel if followed by a back one, as in nigon^ jolega^ 
unless, of course, the preceding vowel is a mutation as in 
wreg{e)an * accuse ' from *wrdgjan. So also g was apparently 
back in such words as belgan^ where a vowellike comes 
between. 

666. Open g and g are unvoiced and written h in the later 
language before voiceless conss. and buzzes, which latter are 
themselves unvoiced (5124) : stihsty stlhj) from stigan. 

656. g, g after front vowels are dropped before the voiced 
conss. /, «, ^ in IWS, as in Upian^ Hnan^ sade^ Ifde = older 
tigfiiauy rignan^ sagde^ Ifgde. YP has rlnan^ where the contrac- 
tion of (ij) into I was almost inevitable, but otherwise the 
non-WS dialects keep the ^, even IKt having tneiden (= ^mai- 
den) = IWS maden^ eWS magden. The IWS contraction after 
back vowel, as in broden from brogden, seems to be due to the 
analogy of the present bredan from bregdan, g is often dropped 
in the combination -t^, as in stiweard from stigweard^ ani, anie, 
and medial ige often becomes ^, especially in later WS, as in 
lifi = older ligep * lies.' 

p, b. 

667. b occurs only initially, and uninitially in the com- 
bination mbj as in lamb (cp ng) and doubled, bb = Qmc bjifj, as 
in ic^bb from *wabja Sk vabi-), h^bban = Goth, hajjan through 
*h{fjan. 

For £ as a graphic substitute for/ see 515. 



150 HISTORY OF ENGLISH SOUNDS. 



SCANDINAVIAN. 

658. The Scandinavian (Scand.) languages fall into two 
main groups : 

(i) East-Scandinavian (EScand.), comprising Swedish (Sw) 
and Danish (Dan.). 

(2) West-Scandinavian (WScand.), comprising Norwegian 
and Icelandic. 

These languages are best represented by the Olcel. of the 
13th century, which, with some exceptions, practically repre- 
sents the parent Scand. 

OttTHOGRAPHY. 

668. The Icel. alphabet was the Latin, as learnt from the 
English. It included, therefore,^ and the less frequent ^. It 
added the new letter p, formed on the analogy of ^, for which 
(B was also written. Length was marked by the British {^). 

VOWELS. 

660. a, 9 i e u o 

ei, 9U (au) ja, jg 

§» 9 ; 97 (ey) y o 



a, 9 (a) I e u 



se y oe. 

661. Of these sounds ei (=fi) corresponds to Gmc ai, as in 
8lein * stone,* and Jo, Ju to Gmc eii, as in ijosa * choose,' 9;ui 
'sick,' gn (which in Icel. is diverged into au) to Gmc aw, as in 
dgup * dead,' a to Gmc ^, as in rap * advice,* = OE stdn, ceomn^ 
€eoc, deady rad (nonWS red). 

662. € followed by older a becomes ja, when followed by 
older u, v it becomes jg, through *ea, *eg [*ia, ^ig) by stress- 
shifting as in Kt (443) and hardening of the first element to 
a cons. : gjgf, gen. gjafar^O^ gefu, gefe. 



SCANDINAVIAN. I 5 I 

668. The i-mutations are nearly the same as in OE : 

a (9). . .^ : mann ' man/ pi mfnn. 

e (ja, J9)...i: fijgld (=:*9teldu) * shield/ pi skUdir. 

u (o). . .y : full ' full/ fylla ' fiU/ 

o. . .0 : koma * come/ komr * comes.' 

a...se: m^^' speech/ mala ^ speak.' 

u...y : briin * eyebrow/ pi brynn, 

0. . .oe : fdr * went/ fosra * bring.' 

9u...9y : /p?^* 'loose/ Igysa 'loosen '= later lam^ lepa. 

ju (jo)...y : {/w^ * sick/ *yi» * sickness.' 
i-mutation is also caused by r=Gmc z (315), as in gyra 
* ear/ from Qmc *auzd. 
564. There is also a u- or ft'-mutation : 

a. . .9 : hgnd ' hand ' = Goth, handu, 

a.«.9: ma/ ' speech/ pi mgl (later ma/). 

§••.9: ffgra ^ do* = OK fffnoan, 
665. Final stressed vowels were lengthened, as in /i2 'thou.' 
Vowels were lengthened before I + certain conss., as in 
Mlf * half.* Consonant length was strictly observed even after 
long vowels and diphthongs, as injfrall nom. 'serf/ 9JCiQ,J)rieL 

CONSONANTS. 

566. The consonants were : — 

BACK FBONT POINT FOREW. LIP 

h - - hr J^, 8 f, hv 

- - hi 

k - t p 

- ■ - hn - 



g J r J? f, V 

- - 1 

g - d b 

n(g) - n m 

t9 was expressed by r^ as in i^r/=OE b^Ut, 
567. What has been said of the OE hr, hl^ hw, hn applies also 
to the corresponding Icel. sounds in such words as hring, hlapa 
( = 0E hladan), hvaty hniga, which in Mnlcel. are pronounced 



152 HISTOEY Of ENGLISH SOUNDS. 

with 01, C3, o, 1 resp. Uninitial Gmc A was dropt every- 
where, as in *a, ddUir=:0^ffeseaA,doMer, except that ^ became 
CD (=^*), as in vaxa =0T& weaxan. The original Scand. h is pre- 
served in ME loanwords such as ilaAter =iIceL sldtr. 

568. Initial Gmc J was dropt everywhere, as in af =Goth. 
Jer, the existing initial Js (as in^V^=OE eorpe) being all 
diphthongic. 

669. Initial Gmc w was dropt before and u, as in iinnin 
( = 0E ffenmnnen), ptc of vinna, also before r and I, where it is 
kept in Norw. and EScand. : (v)reip=OE icrdp^ {y)rang * wrong.' 

670. } and / were voiceless initially and in combinations 
such t&ft (often written pt^ as in lopt *air*), voiced elsewhere, 
as in v'lp^ ^£;5'a=Groth, hajjan * raise.' 

671. Uninitial g after a vowel or vowellike was open,=€, <d 
as in saga * story,' borg * city,* s^gir ' says,' except when doubled, 
and in the combination ng = da, as in l^gg * thigh,' lang * long.' Final 
g became A, and was dropt, as in rfr^=OE drog^ prt of draga, 

572. So also older d and b became voiced J) and / after 
vowels and voweUikes, as in r<^=OE r<§rf, ^^a=OHG geban^ 
except when doubled, dd^ bb, and in the combinations Id, nd, 
mby as in ialda. 

573. k and g were front modified (q\ €ii\) before all front 
vowels, as in i^na 'know,* gpra, and before Gmc^, which was 
preserved in writing, though only as a mark of fronting, Gmc 
gj becoming ^^', as in soskja^O^ scecan, ligg;a=0^ licgan. 

674. Final voiced stops were unvoiced in such forms as gait 
= OE geald, prt of gjalda = OE geldan * requite.' 

676. There were various cons.-assimilations. If, np became 
//, nn, as in gM {gull) *gold,' annar * other,' --Goth, gulp, anpar. 
Gmc zn, zdy zr became nn, dd, rr, as in rann * house* = Goth, razn, 
hodd=^Go\h.. huzd, OE hord * treasure,' v^rri 'worse.' nk, nt, mp 
became hk, it, pp, as in drukkin 'drunk,' bait 'bound,* prt of 
binda (through *band, *bant), kappi 'champion' (cp OE cpnpa), 
the original nk etc being often preserved in EScand. 

676. n was dropt finally in monosyllables and endings, as in 
d, e = OE on (Goth, ana), in, finna = OEJindan infin. ; and before r, 
1, 9 and elsewhere, with lengthening of the preceding vowel, as 
in Par ' Thor ' = OHG donar, OE punor, gds=:OB.Q gam ' goose.' 



SCANDINAVIAN. 1 5 3 



INFLUENCE ON ENGLISH. 



677. The earliest Scand. invaders of England were mostly 
Norwegians, who were followed by Danes, all Scandinavians 
being included under the term 'Dane' in OE. Danish and E. 
were spoken side by side in England for many centuries with- 
out much influence — at least of Danish on E., even the INorth. 
texts showing no traces of it. In the i ith cent. Danish words, 
such as lagu *law* (Icel. Igg pi from *lggu)^ ceaUian (Icel. 
kalla) had penetrated even into WS, and in the 13th cent, 
their number largely increases, not only such words as gerseme^ 
fvontreape =Ice\, gprsimi * treasure,' vandrafoi * difficulty,* but also 
grammatical words, such as bg>fie * both * = Icel. bdfoi (OE bd, begen) 
being firmly established in the Southern dialect. The Scand. 
element is, of course, stronger still in the East-Midland Ormu- 
lum of the 13th cent. Such words as summ * as,' bqpe * booth ' 
in the Ormulum are distinctly EScand. (Danish) as opposed to 
WScand. (Icel. serny buf), 

678. The Scand. words in OE sometimes preserve g in the 
form of 0, as in hold * yoeman ' from hgld^ hgjp (cognate with 
OE halep ' hero '), where the is partly due to the analogy of 
the OE adj. hold * faithful.' In other words, such as lagu^ g is 
unrounded, as in the IWS mann etc. Of the diphthongs, ei is 
preserved unchanged, being expressed sometimes by W, but 
generally by eg^ ag^ as in scegP^loi^. skeip 'war-ship'; and 
gy was probably levelled under it. ou becomes cf (=p ?), as in 
ora (a coin) = Icel. aurar (pi) from Lt aureus. It is remarkable 
that (?, often appear as t^, « in OE, as in £/m, \)ur= Orm^ por^ 
whence our Thursday, which cannot be explained from the 
original OE punresdteg. This change is explained by the pre- 
sent Dan-Sw-Norw. pronunciation of dose 0, both long and 
short, as }) — a sound between and u. This Scand. was 
afterwards levelled under the E. sound, so that we find Orm, 
blame (Icel. blomt) in ME. 



154 HISTORY OF ENGLISH SOUNDS. 



MIDDLE ENGLISH SOUNDS. 

DIALECTS AND TEXTS. 

679. The ME dialects are mainly continuations of the cor- 
responding OE ones, but it is convenient to designate them in 
some cases by different names. The four main divisions are: 
Northern or Northumbrian (North.), Midland (Ml), correspond- 
ing to the older Mercian, Southern (Sth), and Kentish (Kt). 
Sth and Et are included under the common designation 
' Southern EngUsh* (SthE). Ml is subdivided into West-Mid- 
land (WMl) and East-Midland (EMI), and these, again, into 
NoHh-west-Midland (N WMl), South-west-Midland (SWMl), N(nih- 
east-Midland (NEMl), and South-east-Midland (SEMI). A special 
subdivision of Sth is South-Western (SthW). It is to be noted 
that though Sth represents geographically the old WS it also 
shows strong Ml influence. This mixture of dialects is still 
stronger in the later language of Ch. 

680. It is impossible to draw any absolutely definite line 
between ME and OE on the one side and MnE on the other, 
but, roughly speaking, fully developed ME may be said to 
extend from 1150 to 1450, the period between laoo and 1400 
being especially well marked and well represented by written 
documents. The pei-iod from 1050 to 1150 may be distin- 
guished as Old Transition (OTr), that from 1450 to 1500 as 
Middle Transition (MlTr). The diflBculty of drawing a line is 
increased by the varying speed of change of the different 
dialects. The most conservative dialects were the Southern, 
especially Kt, the most rapid in their changes the Northern : 
the eNorth. dialect of the 13th cent, is, indeed, almost on a 
level with eMnE. Taking' the SthE dialects as the standard 
we may call everything before 1300 early Middle English {eilLE)^ 
everything after 1300 late Middle English (IME). 

681. If we take SthE as the standard, we may define OE as 
the period of full endings (nwna^ sunne^ snnu, stdnas), ME as the 
period of levelled endings (mone, sunne, sune^ stgnes) — weak vowels 



MIDDLE ENGLISH: DIALECTS AND TEXTS. 1 55 

being reduced to a uniform e (=1?) — ^ MnE as the period of 
lost endings (mootif sun, son). 

682. The most important of the OTr texts is the latter part 
of the Laud ms of the Chronicle (Ld), which was written at 
Peterborough between 11 24 and 1154, and belongs therefore 
to EMI : it shows a mixture of literary WS and Ml forms. 
The older ms of Layamon's Brut (Lay.) was written before 
1200 in a WMl dialect, and its mixture of OE and ME 
forms classes it with the OTr texts. Many 12th cent, texts, 
such as the Hatton ms of the Gospels, Morris's Old-English 
Homilies (Horn.), show a mixture of OE and ME forms which 
is the result of copying from OE originals, and only partially 
modernizing them: such texts do not represent any actual 
language. 

683. The Ormulum (O.), although written probably before 
1200, shows a fully developed and well defined ME dialect — 
probably EMI — preserved in an autograph ms of the author 
in a rigorously consistent phonetic orthography, which makes 
it the standard text for ME generally. The other chief eEMl 
texts are the Bestiary (Best.) and Genesis and Exodus (GE). 
Havelok (Hv) is, like most of the popular poems, preserved 
only in mss showing a purely scribal mixture of different 
dialects and periods, and which cannot, therefore, be quoted 
to show the dialect of the original, except when the form in 
question is borne out by the rhymes. eWMl is represented 
by the second text of Lay (Lay.^). The poems in the Harleian 
ms 2253 (Harl.), written in Hereford about 1307, may also be 
considered eWMl. IWMl is represented by Piers Ploughman 
(PPl), and, in its latest stage, by the poems of Audelay (Aud.), 
written in Shropshire in 1426, the ms being probably the 
author's autograph. eNWMl (Lancashire) is represented in 
the Alliterative Poems (Pearl, Cleanness, Patience) edited by 
Morris (AllP). The later EMI is well represented by Bobert of 
Brunne's Chronicle (RBC), Brunne being in Lincoln. 

684. One of the earliest North, texts is the Metrical Psalter 
(Ps), but the ms is later. The Cursor Mundi (CM) and Me- 
trical Homilies (MH) are early 13th cent. Then follow the 
Prick of Conscience (PC). The ms of the Yorkshire Townley 



156 HISTORY OF ENGLISH SOUNDS. 

Mysteries (TM) was written about 1450, but the rhymes show 
an older language. 

685. The Sth dialect is represented in its earliest form by 
the lives of St. Katherine (Kath.), Juliana (Jul.) and some 
allied pieces, also by the unpublished Cambridge (Coi'pus) ms 
of the Ancren Biwle (AB^), although this Corpus ms seems to 
show Ml influence ; the forms common to this group of texts 
may be distinguished as 'earliest Sth.* Pure eSth is best 
represented by Morton's text of the AE, the more western 
dialect of Gloucestershire by Robert of Gloucester's Chronicle 
(RG), which, however, belongs almost to ISth. 

586. eKt is represented by the Kentish Sermons (KS), IKt 
by the Ayenbite of Inwyt (Ay.) in a ms of 1340, written by 
the author himself in a very pure and consistent dialect. 

587. ^he first beginnings of a common literary dialect are 
seen in the works of Wiclif ( Wicl.) and Chaucer (Ch). Wiclif 
was a native of Yorkshire, Chaucer of London. Chaucer's 
rhymes show a considerable fluctuation between EMI, Sth, 
and Kt, but the basis is Ml. 



ORTHOGRAPHY. 

588. While the linguistic change of OE into ME is so 
gradual that it is difficult to tell where the one ends and the 
other begins, the orthographic change is abrupt and complete : 
it amounts, indeed, to the introduction of a totally new basis 
— the eNorman Fr orthography, modified, of course, in detail 
by the traditional British orthography. 

588. For some time after the Norman conquest in 1066 the 
two orthographies continued to be used side by side without 
influencing one another to any great extent, just as the lan- 
guages themselves were kept apart. The influence of French 
language, writing, and orthography had, however, begun to 
show themselves even before the conquest. The feeble reign 
of Edward the Confessor was, indeed, in its tame submission 
to Norman influence, nothing but a preparation for the com- 
pleter conquest that was to follow. The influence of the 
French language is shown by the appearance of such words as 



MIDDLE ENGLISH: ORTHOGRAPHY. 1 57 

9oU 'hebes', capun * gallinaceus ' in early nth cent, glossaries. 
The influence of the French handwriting has been described 
already {222), The influence of Fr orthography is seen in such 
spellings as euen for efen in i ith cent. mss. 

690. This influence was at first purely Norman. The ac- * 
cession of Henry II of Anjou in 1 154 brought in the influence 
of other dialects, and the loss of Normandy in 1 204 paved the 
way for the influence of literary Parisian orthography both in 
its earlier and later form. 

691. When the popular Latin of Gaul was written down — 
which was probably not much earlier than the ninth cent. — 
its sounds were represented by their nearest symbols in the 
contemporary Latin alphabet. But by this time the tradition 
of the classical pronunciation — stiU preserved in the Celtic- 
English orthography — had been partially lost. The diph- 
thongs a and (e had been levelled under simple e^ and y had 
come to be a mere variant of i. So when Lt u was fronted to 

f in OFr, as in hne^ the old u was kept a^ the symbol of the \^ 
new sound. Meanwhile Lt u and had become }>— a souni 
between (u) and (0)— which was at first written indifierently u 
or 0, as in gule^ gde^ curt, cort from Lt gnlam^ *cdrtem (from 
coAorlem), the soon becoming general, and thus being confused 
with the open from Lt 8, au, as in port s^jjortum, chose ^causam. 
The eOFr diphthong (ou), as in douz (earlier ddz) from Lt dulcem, 
was smoothed into (uu) in IParisian, and so ou came to be the 
symbol of (u, uu) instead of the older t^^ 0, as in gouley court. 

692. Li the conss. Lt c before front vowels became first q 
and then (tj), which was the Ficard pronunciation in such a 
word as del from Lt caelum. In the other dialects this c became 
(ts), and then simple (s), as in the present Fr. cA was at first 
used to denote (k) before front vowels as in Italian : chi from 
qvt; but afterwards became the regular symbol for the (tJ) 
which in Parisian developed out of c followed by a (and in 
other cases), as in ciien from canem through ^Qil[. Lt g 
was fronted under the same circumstances, and became (d^), as 
in gesle from gesla, which remained through the OFr period, 
becoming (5) in MnFr. Lt j was stopped into m, which then 
became (ds) being written i otj\ as in^a from Lt Jam, The 



158 HISTORY OF ENGLISH SOUNDS. 

combination (kw) was expressed by the Lt qu, as in quel from 
qvdlem^ the (w) being soon dropt in pronunciation. So also the 
(gw) in such a word as langne from lingvam lost its (w) in IPar. 
Hence qu and gu came in lOFr to be regarded as symbols of 
* hard ' c and g resp. The new ligature w was formed in North 
Fr to express the Gm (w), as in warde, for which the other 
dialects have (gw) ; guarde, later garde. The Lt v itself had 
long ago lost its (w)-sound, and had come to represent that of 
the voiced sound of/. Lt z still kept its original value (dz) 
when initial or medial, becoming (ts) when final, as in assez — 
our assets — from Lt adsatis ; it was not till the lOFr period 
that these compounds were simplified to (2) and (s). Hence 
the (z)-sound which Lt * took between vowels was expressed 
by the traditional «, as in rose from rosam. Already in eOFr 
the earlier (z) before voiced conss. had been dropt with length- 
ening of the preceding vowel, as in isle from Lt Insulam, Hence 
in lOFr s is sometimes inserted as a mark of length, as in 
pasle ioT pal{l)e from lA, pallidus, 

593. It will be seen that OFr orthography was phonetic on 
an unphonetic basis (228) : it is not till the close of the OFr 
period that * etymological * spellings begin to crop up. The 
only exception is the writing of silent initial Lt h in such words 
as hum from hornd^ but as this is only done when a vowel pre- 
cedes — * the man ' being written liim — it is probable that the h 
was meant to indicate the hiatus, and was, therefore, partially 
phonetic. 

694. The basis of ME orthography is, as already remarked, 
French modified by the OE tradition, the OE elements being 
gradually eliminated more and more. Conversely, however, 
some of the earliest ME texts show a basis which is still 
mainly OE, only slightly modified by Fr. This is especially 
the case with the Midland 0. as compared with the Sth texts 
of the same period. 

595. The influence of Fr is most strikingly shown — as 
far as the vowels are concerned — ^in the substitution of u for 
OE ^y, y in WMl and Sth, as in suime^ y«r=OE *y»», fyr. 
The long sound was sometimes written vi, as in huiren * hire ' 
(AR) = OE h^rafi, in IME wy, which is frequent in PPL In 



MIDDLE ENGLISH : OBTHOGBAPHY. 1 59 

OFr ni had generally the value ('yi), as in fruit ; this diph- 
thong was smoothed to (yy) in the E. pronunciation of Fr, 
and was consequently employed to represent that sound. The 
use of for short (u) is later than that of u for (y). It does 
not occur in AR, and does not become general till the end of 
the 13th cent. Remarkably enough, there are several instances 
of it in Lay. : wonede^ icomene^ wode. It is fully established in 
Lay.^. The tendency is to write {or u in proximity with 
letters that resemble u in shape, especially u (=f), », m, w. 
Initial u was, however, often written r, which was freely as- 
sociated with n etc, as in vtuler, in IME is also generally 
written instead of u when followed by a single cons, and a 
vowel, as in lote * but ', cordge * courage ', for, as Fr. (y) 
was much more frequent than the (o)-sounds in this 
position, lute would have suggested (byyto). The use- of 
IParisian ou to express (uu), as in hous— 0£ kus, became general 
in 1M£. This au also occurs in Lay., as in out {vt in the 
second text), wi^oiiten, conjie^ where it cannot be of Parisian 
origin. But the eFr diphthongal ou had in many cases the 
sound of }-)3:, which is so near (uu) as to make its symbol a 
very natural one for the latter sound. The desire to get a 
new symbol for (uu) would, of course, assert itself as soon as 
u had become general in the value of (yy) — that is, from the 
very beginning of ME. 

696. The OE y was, as we have seen, completely superseded 
by « in the Sout£ In OTrMl (and probably in North.) it was 
unrounded into i. Ld still preserves the old y in byrien^ 
myMter^ but. these words are also spelt with i ; we also find in 
Ld such spellings as «««^,^r=OE synna^ fyr, y is rarely 
written for i in Ld, oftener for t , as in 9cyr adj. and sbst, tyma ; 
•probably y was regarded as a ligature of i and j. In O. y 
disappears as completely as in the South, except in foreign 
words. In IME y was revived as a variant of i in proximity 
with n, tfiy u, «7, in order to avoid confusions of form, as in 
byndetij wyues {^^tcives), which confusions were often avoided 
by writing initial i as a capital : Inne =ynne. In Ch there 
is a tendency to write y for i, as in Ld. 

697. OE a was kept in O., but was necessarily confined to 



\ 



1 60 HISTORY OF ENGLISH SOUNDS. 

the long sound, the short as having become a. In the South 
Fr influence caused its disuse. Here the OE a was expressed 
by ^, as in efter=0^ isfier. As OE ea was smoothed to x* in 
ME, ea was used as the symbol of the latter sound both when 
it corresponded to OE ea,as in deap^zsA when it corresponded 
to OE ^, as in learen=-Qi^ laran {da}^ larenn in O.). Even in 
eME e is frequently written for the open as well as the dose 
sound, and in IME such spellings as defi^ leren become uni- 
versal. In IME the close e is sometimes written ie, ye^ both 
in Fr words, such as meschief^ and in £., as in /i^(Ch)=OE 
leof. This spelling is the result of the Anglo-French smooth- 
ing of OFr ie (i-ee) into (ee). 

698. A distinction between close and open long is only 
exceptionally made in eME (as in the AR) by writing the 
latter oa in such words as moare from OE rn^re^ the oa being a 
natural compromise between the older a (still preserved in the 
earliest Sth) and the later 0. 

699. It will be observed that the digraphs ea, ie, oa, mi, ui 
are strictly confined to long vowels, except in some of the 
earliest texts. 

600. The OE^ and 9 are both preserved in Lay. — ^where / 
is generally written initially, <3f non-initially, as in IWS — and 
in AR, where they are distributed more at random. The EMI 
Best, and GE are remarkable for writing "8 everywhere, while 
O., which belongs to the same dialect, has only J), p entirely 
supersedes d in IME, being itself gradually supplanted by th — 
probably brought in by Fr scribes who occasionally employed 
it in learned Latin words. Isolated th% in native E. words 
occur very early — even in the OE period — and the transposed 
ht occurs in Harl. (^(?/4^=0E Up), and frequently in MH {wiht 
prp), and elsewhere. Other combinations occur : dh (KS), dh 
(GE), hd (GE), dp, dd. It may be noted that p has survived 
almost to the present day in the contraction ^=>&^. 

601. The OE rune-?^7, which is still used in 0. and AB, was 
soon superseded by the Fr ligature. The pronunciation of 
OE ow %\/tdA ou led in IME to the general use of t^; as well as 
u as the second element of graphic as well as phonetic diph- 
thongs, as in how:=.QYA hu. Conversely, %o was sometimes used 



MIDDLE ENGLISH: ORTHOGRAPHY. l6l 

aa an abbreviation of lou, as in wde, w after a cons, was often 
written u : ^uerd^ huo (Ay.). 

602. In O. / is still used for intervocalic (v), as in lufenn^ 
OE Infian^ and this usage was more or less kept up in North, 
to the end of the IME period, even TM showing instances of 
it. But O. has in a Fr word aervenn once instead of his usual 
ierfenn. Medial v is regular in AB, as in Aeovene, and is very 
frequent initially, as in vorj) by the side oiforj). /is always 
preserved finally, as in ^, to prevent confusion with the vowel 
ui *liu, for instance, would suggest a diphthong. For the 
same reason / is written before voiced conss., as in Aefde:= 
0£ Aafde, ph is only found occasionally in learned words 
taken from OFr. 

608. s is generally written for the voiced as well as the 
breath sound. 99 \r sometimes written 9c, which had the sound 
of (ss) or (s) in OFr: thus AB has le9cun 'lesson,' ble9cen 
' bless,' and GE and CM have bli9eed ' blessed.' The lOFr z 
is used pretty regularly in the Ay. for voiced 9, as in zigge^ 
aze. Elsewhere it is rare in E. words. Thus Ch has Pize : 
rUe, such spellings as tcezele being exceptional. In eME z has 
the older value (ts) in Fr words, especially in the combination 
nz, such spellings as the plurals be^zannz (O.), ve9tiMenz (AR) 
lasting down to Ch. 

604. The general disuse of the OE c before e and % is the 
result of its double pronunciation in Fr. O. always writes k 
before e and t, often also before a, but alternating here with c, 
which is always written before 0^ u and cons& : kepenn^ king^ 
kare, care, com^ eumenny dafi^ cwen. The usage in AR is the 
same, except that it writes k freely before 0, u, and conss. ex- 
cept ia : ku, cunnettj kniht^ cwene. In eME e has its eOFr value 
of (ts) : OE milUe appears in Ld as mUce, in O. as millce. tc 
has the same value in some eME spellings of OE bleUian : Ld 
has bletcady O. blettcedd by the side of bletUedd. AR has the 
half-etymological spelling 9eldeene=0^ 9eld9ene (524), where 
the other text has the phonetic selUene, It is not till IME 
that we find c by itself used for (s) in E. words, as in alee 
* also ' (AllP), waee * was ' : face (AllP). qu for cw is rare in 
AR, but soon becomes general 

M 



1 62 HISTORY OF ENGLISH SOUNDS. 

606. The OE c=d is still kept in Ld, and is even used for 
the Fr ch^ as in Ricard. ch = OE c is rare in Ld, being used 
only as a variant of uninitial h (606), but it is fully esta- 
blished in O. Its doubling is ceh throughout ME, the first 
e being, however, occasionally dropped. Ay. has also chch^ as 
in wychche by the side of wyche, usually wicche, 

600. A=c, o is kept in eME : thus O. has heh^ Jmrrh^ nihkt, 
eh occurs occasionally, as in Butch (Ld), pirch, almichii (KS), 
unricht (CM) ; it was no doubt disused to avoid confusion with 
the Fr ch. But scribes used to the Fr spelling evidently felt 
that h was not a suitable sign for so strong a cons. OE writes 
^, as in 9og^ rigt, and occasionally c before ^, as in broete. Hv 
writes eth for hty as in ricth. Other mss write ^, which is 
common in IME: thus Wicl. has hi^, nov^L North, has ^,^ in 
the earlier texts, gh in the later: thus CM has no^ {noghf). 
This gh gradually spread south, and is fully established in Ch. 

607. For OE #d, which is kept in Ld, 9ch is written in eSth, 
as in ^Aeome, achrifi (AR), of which the sh of O. (occurring oc* 
casionally in eSth also), as in shame^ shriffte^ is probably an 
abbreviation. Ay. writes regularly m, as in strive^ vlest. This 
99 is probably an abbreviation of the ssc of ES, in which the 
doubling of the s is probably meant to indicate that the w 
stands for a simple sound. ES has the various spellings 
flesce, fleece, ssipe, 9rifte, 9=^9h occurs also in the later text of 
Lay., and elsewhere. Barer spellings are she and eh, which 
latter occurs in ON (ehadde prt) and often in Aud. by the 
side of sch, as in ehame, schamyd. IME varies between seh and 
sh. The doubKng of th is written tsh in O. as in ennglissh, 
sehsy 9h9, ash in AB. 

608. Li ME the difference in form between the English g 
and French g was utilized phonetically, the former being 
assigned to the open sounds e, co, the latter to the stop a and to 
the French soft g and the ME development of OE g in eg, ng, 
which had nearly, if not quite, the same sound (737). Orm 
uses g for open y, as in gung, manig, gh for open g, as in 

foUghenn — where the earliest Sth has A, as in/Men'^g for the 
stop, as in god=0^ god. The initial soft y in Fr words is 
sometimes written g, but oftener j in the AB, as in gugemefd. 



MIDDLE ENGLISH: STRESS. 1 63 

j^9^» 99 18 ambiguous in ME, as it may be either OE or Scand. 
99 (^^99^) or c^ (Sth seggen). In Ld initial 9 is often written i, 
as in iaf by the side of 9eaf. This spelling was no doubt 
given up in order to avoid confusion with the Fr / Best, 
and GE are remarkable for their general use of ^ to the ex- 
clusion of ^, as in 9in9y gimg * young/ The oldest text of CM 
has 9i = ^f as in giet ' yet/ gieme = OE geman, the later 
North, texts writing yA (PC) and y : yMt, yit. IME varies be- 
tween g and y : gong^ yong. This ME consonantal y is probably 
due to the lOFr writing of j^ initially instead of «', y being as 
rarely used in eOFr as in eME. It is also possible that the 
consonantal use of y may have been suggested by the lOE 
spellings yorjie etc (434), where the y practically denoted ( j). 
The use of gu to express hard 9, both in Fr words, such as 
guard, and E., such as guest, guilt, is due to later OFr, in 
which the older (w) in langtie, guarde had become silent. This 
older pronunciation was never introduced into England, be- 
cause most of the ^-words were pronounced with w {loarde 
etc) in ONorman. 

METBE AND STBESS. 

609. The earliest ME verse, as seen in Lay.^, is a con- 
tinuation of a metrical revolution which began in the OE 
period, and went hand-in-hand with the decay of the old laws 
of alliteration and the gradual development of rhyme. Laya- 
mon's four-stress metre agrees with the old alliterative metre 
in the freedom with which unstrest syllables are 'omitted or 
added between the stresses and before the first stress, and 
in being based on a compromise between the natural stresses 
and syllable-quantities of the language ; but while in the old 
metre the natural stress is the leading element, to which the 
quantity is always subordinated, the contrary is often the case 
in the ME four-stress metre, in which syllables that are quite 
stressless in ordinary speech can in verse take the full stress 
required by the metre. While a dissyllable like sune, with the 
stress-syllable short, has only the metrical value of a mono- 

^ M. Trantmann : Ueber d«n yen LaSftmon*B (AngHa, ii. 15$). 

M 7. 



1 64 HISTORY OP ENGLISH SOUNDS. 

syllable, as in the old metre, a similar word with the first 
syllable long, such as sunney is allowed to take a metrical 
stress on each syllable, as is shown not only by the structure 
of the verse, but also by rhymes such as toes (=u?as) : londes. In 
such cases as these — which are especially frequent at the end 
of the line — the long syllable took an extra prolongation, 
so as to fill up part of the time of the following one. The 
following are types of this metre as employed by Lay, : 

]?a 'c5m him *to an 'hende 'cniht. 

bi 'US he *sende "word •)?§. 

•Ar)7ur 'is J^e 'kenneste^ "mon. 

•ofte "wes J?e 'drake 'buven, 

mid -seolvre -and mid 'gol'de. 

*]7eines 'wunder -bli-J^e. 
This metre is identical with the OHG one of Otfned, which 
was based on the late Latin hymn metres. 

610. The first to employ a strictly syllabic metre was Orm. 
His metre consists of pairs of half-verses, the first having 
eight, the second seven syllables with a regular alternation of 
strong and weak stress, the first half-verse beginning with a 
weak and ending with a strong stress, the second beginning 
and ending with a weak stress, the last syllable but one of 
the second half- verse being always long : 

annd 'broj^err 'min i 'Godess '.htis 
jet "6 )7e •J?rid[d]e 'wise. 
Such a word BiAfaderr never occurs at the end of a line in O. 

611. This metre was probably originally simply a doubling 
of the olde^ four-stress line, the regular alternation of weak 
and strong stress afterwards depriving the last syllable of the 
second half- verse of its metrical stress. 

612. As in the later poetry, weak final e is elided before a 
vowel or A, as mforr lufe off Crist. 

618. In the interior of a verse naturally weak syllables 
often take the metrical stress, in order to secure the regular 
alternation of strong and weak stress, as in 

aff'terr )>e 'fiseshess *kinde. 

614. In IME the lengthening of the vowels in such words 

^ nkSneste. 



MIDDLE ENGLISH: QUANTITY. 1 65 

as ndme^OE nama led to the abandonment of the earlier 
quantitative restrictions. But Chaucer still retains part of 
Orm's freedom of stress. He throws forward the stress to 
any naturally unstrest syllable containing a full vowel, as in 
mdJryng^ bod'y, whtl'om, and, of course, on to the half-strest 
second element of a compound, as brim'slpn. But he never 
throws the stress on to the weak e of such words as qfter^ 
name. Throwing back of the sti'ess, as in 'uncouj) for the usual 
ufi'coup is, on the other hand, rare. 

616. Chaucer's metre seems to show that the nominal, pre- 
fixes cd'f miS", an-, which took the stress in OE, had now 
thrown it forward, as in al'mighty^ mis^deed^ uwreste^ un'fynde, 
bp- still keeps its stress in biword, but elsewhere it loses its 
stress, probably by analogy of the stressless be- in beginnen 
etc, as in bvh^§9t€. The old separable compound verbs also 
throw their stress forward by the analogy of the inseparable 
forgiven etc, as in up'rue^ out-ride. 

QUANTITY- 

Besides the indirect evidence of the metre and the laws 
of sound-change, ME quantity is in many cases determined 
by the direct evidence of the spelling. 

616. The most thoroughgoing attempt to mark the quantity 
consistently is that of O. In this text every consonant that 
is final or followed by another cons, is doubled after a short 
vowel, as mfatt^ crisatenndam^ }'»»=0E f»», in. This shows 
that the OE distinction between in and inn must have been 
lost in pronunciation as well as in writing, so that all final 
conss. were lengthened after a short vowel, as in MnE. Such 
spellings as ic amm, scipp = eom^ 9cip occur already in Du., and 
many examples might be quoted from later texts. Ld has 
namm prt, ieit * yet ' etc ; godd occurs in Lay., Jul., GE, CM. 
The opposite tendency is to shorten long conss. after a long 
vowel ; and so double final conss. came to be associated with 
preceding short vowels, and single conss. with long ones in 
writing as well as in speech. There are isolated traces in OE 
of Orm's doubling of a cons, before another cons, in such 



1 66 HIBTOET OP ENGLISH SOUNDS. 

spellings as follce^ illce (Smith's Bede), where, however, the 
doubling may really indicate cons, length. In efter (Jul., Ay.) 
the doubling of the / may mean voicelessness. That Orm's 
doubling did not mean cons, length, but had been reduced to 
an abstract symbolization of vowel-length, as in MnE and 
Om, is clear from his extension of it to unstrest syllables, as 
in broferr^ for already in OE double conss. are shortened 
after an unstrest vowel, whether long or short (409), addi- 
tional ME examples being afforded by such spellings as 
leofmones ' lemman's ' = OE leofmannes, meinfule = OE magenn 
fulla in Jul., 9unfule = stfnnfulle in AR. Where the cons, is 
followed by a vowel, as in sune^OEt sunu^ it was not possible to 
double the cons., because it would then have been pronounced 
double, and stme would have been confounded with wnnc 
' sun.' That sunne was really pronounced with long or double 
n is proved by the metre, which allows sunne to come at the 
end of the verse, and rigorously excludes sune from that 
position. 

617. Here, then, Orm's clumsy spelling breaks down com- 
pletely, and he feels this himself, for he often marks the 
shortness of the vowel in such words as stme with a {^), as in 
t&ienn, n&me, cAele:^OIcel. taia, OE nama, e§le. Often, too, he 
marks length with the old accent, as in Idre = OE /ir, which 
he often doubles, or even trebles, especially before ^, as in ut. 

618. The old accent is rare elsewhere in eME. It occurs, 
for instance, in the Proverbs of Alfred (Oxford ms), as in 
ilSredy dSmen^ sSe, ES has h66t = hate^^ and the anomalous 
uerSS^OYifyre dat. 

619. The OE doubling of long vowels occurs sporadically 
in eME, and becomes very common in IME, first in North, 
and then in Sth, being strongly developed in WicL and. Ch. 
The doubling is carried out most regularly in monosyllables 
such as itoofiy deed = OE %tdn^ disd, dead, but also medially, as 
in keene, oother. Doubling before cons.-groups is frequent, as 
in hoard, foond = OE hord, fand. i and u are hardly ever 
doubled in IME, to avoid graphic confusion. 

Quantity is also indicated by the use of the digraphs ea etc 

(599)- 



MIDDLE ENGLISH: QUANTITY. 1 67 

620. Orm's doubling of final conss. is found in other ME 
texts as well, but only occasionally, single final consonants 
being the rule, as in a/, man^OE eall, mann. 

621. In eME short vowels, as we have already seen (616), 
retained their original quantity before single conss., the cons, 
beiiig lengthened finally in a stressed syllable, OE nama, in, 
inn appearing in eME as nam€,in{n)y in{n)y the OE double conss. 
before vowels being kept, as in ^nne, 

622. In IME short vowels before single conss. followed by 
a vowel were lengthened (* new-longs '), name becoming ndme^ 
the combination short vowel + double cons, being kept, as in 
aunne, which even in Ch is not allowed to rhyme on sone = 
OE mnu. The lengthening in name etc is proved not only 
by the evidence of MnE, but also by doublings, as in bjfoorey 
and rhymes such as Aope : grope =^0'E kopian, grdpian. 

628. The high vowels i (and OE y), u are never length- 
ened : writen ptc, dide, 9une = OE writen, dyde, 9unu, except, of 
course, where they had been lowered before the IME period, 
as in evel=0^ jfel. 

624. Short vowels in final stressed syllables (which were 
generally monosyllables) could not be lengthened, because the 
following cons, had been already lengthened in eME : imal, 
pap, stqf, swan, blak, 9ad, glad, sap, troA, god =: OE smal, pap, 
staf, swan, blac, sad, glad, sap, trog, god. So also the preterites 
gaf, spah, Irak, sat, bad kept their vowels short, as proved by 
such rhymes as yaf : staf in Ch, the MnE gave, spate, brake 
being due to the analogy of the long vowels of the infinitive, 
whence also gave got its v ; sat has kept its vowel short be- 
cause the infin. is sitten with a short vowel. 

626. Apparent exceptions to these laws are mainly due to 
the ME form being taken not from the OE base, but from 
some oblique case or derivative. Thus the short vowels in 
nance, falwe, geltoe point to the OE inflected nearw{e)y fealw-, 
geolw", OE ealu has gen. ahp, not *ealwes, and hence its vowel 
is regularly lengthened in MTC : ale. bale and mele from OE 
bealu, melu, gen. bealwes, melwes are exceptions ; probably these 
words were but seldom used in an inflected form. 

626. Many OE neuters with short root-syllable take a final 



1 68 HISTORY OF ENGLISH SOUNDS. 

e in ME, thus OE gehed ' prayer ' appears in O. as hede = MnE 
lead, the OE pi gehedu being apparently taken for a mase. or 
fern, sg like medu ' mead/ cam. Hence the long vowels in 
the IME dale^gate {gate\ blade, bede, hole, cole = the OE neuters 
dal, gat, blad, gebed, hoi, col. hoi and col also occur, and O. 
has gocc against MnE i/oke^O'Ei geoc. Many neuters have only 
the short forms : bap, glas, chaf, blah, fat (vat), brop, lot, ^o^=OE 
bap, gUss, caf, bUBc,fat, brop, hlot, cot. 

627. The lengthening in the adj lame is due to the OE 
weak form lama, that in late to the OE adv late (lat being 
the OE adj). 

628. The fluctuation between short and long vowel in wel 
ssid get iB OE (387). 

629. Medial shortness before a single cons, is preserved in 
IME before certain endings (* back-shortening'), but with many 
exceptions, and fluctuation in some words : 

-er : hamer, stameren, water, fader y feter, oter^ coper. Excep- 
tions : aker, taper, over. The true ME compar. later is preserved 
in MnE latter, later being a new formation from late. MnE 
(raaiSer) and vulg. (reiiSor) points to a IME raper and roper 
=0E hrapor. 

-el: shakel, sadel, watel, netde, hovel. Exceptions: navele, 
cradel, mapel (=0E mapulder\ stapel, wesele, evel (from OE t/fel, 
IKt evel). 

-en : neven, rekenen, soden, troden. Exceptions : raven, even, 
henepen, stolen, cloven. The -en of the infin. does not shorten, 
as in bapien, apeken, because of the analogy of the pres. 
forms i epeke etc. 

-ing. hering. 

-i (=0E -ig): mani,peni, bodi,popi. 

630. Some of these back-shortenings may be explained by 
inflectional forms in which the vowel is followed by two 
cons.^ as in Orm's seffne, but to many of the words this does 
not apply. It will be observed that the endings are all 
vowels or vowellike conss., and the real explanation probably 
is that the lengthening was shifted from the root vowel to the 
ending. Final ig in manig was long according to Orm*s 
spelling. Lengthening weak e alone or followed by an un- 



MIDDLE ENGLISH: QUANTITY. 1 69 

vowellike cons, waa less easy and natural, and hence the 
lengthening was thrown on the root-vowel in such words as 
name^ metCy nosey staves, naked^ with few exceptions. 

681. The same influences which preserve vowel -shortness 
in IME sometimes shorten long vowels, as in laper from OE 
leapor through ^IfPer. The shortening in MnE sorry ^ silly, ME 
eni, aff2=0E sdrig^ gesalig, anig may be due to the *i. 

632. Long vowels are regularly shortened in ME before 
two conss., except, of course, before those cons.-groups which 
lengthen short vowels (635). Thus O. has wis * wise ' but 
toissdom ' wisdom,' demenn prt demmde, dradenn ptc forrdredd 
=i01&fordradd, hdll^ 'holy' pi hallghe^ vb hallghenn 'hallow.* 
So also in lasse comp., wesste ^ desert,' adj, blosstme, soffte = OE 
l^ssa, weste, hldstme, sdfte. The shortenings in naddre (neddre O.), 
udder, fodder from OE nadre, uder, fodor are due to the later 
OE doubling : naddre etc (410), whence also the shortening 
in comparatives, as in Chaucer's gretter, derre, depper^ pos. grfl, 
der, dep^O^ greal, deor, deqp. 

633. Length is, however, often kept before st, as in Orm's 
Crist, l€Bst superl., ast (=0E east), pre{o)st, bre(o)st, 

684. Shortenings, and exceptional retention of shortness in 
IME, are sometimes the result of want of stress, as in 1^ 
(uss O.), from which some texts have the emphatic ous =z us, 
MnE Aave (hasv) by the side of behave (bi'heiv). 

636. The OE lengthenings before vowellikes + cons, are 
kept up in ME, but with certain restrictions. The most im- 
portant of these is, that the second cons, must be voiced: 
vowels are never lengthened before nc, nt etc, O. writing 
drannc, drinnkenn, stunnt, sallt, hellpenn etc. Lengthening is 
also regularly barred by back-shortening : thus O. writes aid, 
but allderrmann, elldre cp, cKUd^ but pi chilldre. The -^» of the 
ptc has, however, no back-shortening influence : cp stngenn \ 
findenn with sungenn,fundenn. The length both of the infln. 
and ptc is really due to the influence of the monosyllabic 
preterites /<x»J etc. Those verbs which have no monosyllabic 
prt keep their vowels short throughout : brinngenn (prt brohhte), 
mnndwenn, senndenn, wenndenn, which last sometimes has a 

' The following exAmples are from 0., unless some other reference is given. 



170 HISTORY OP ENGLISH SOUNDS. 

single ft, which may be a mere error, blendenn * blind' has 
a long vowel by the analogy of blind, stanndenn has a short 
vowel because its prt stod, though monosyllabic, does not end 
in nd. The older forms of such a verb as flndenn were, there- 
fore, ^Jinndenn^ ice findcyfand^ ^funndenn. In spite of the length 
in such words ad ende, tunge, there seems to have been some 
tendency to shorten before a following unstrest or half-strest 
syllable independently of back-shortening : winndeclut^ anndstoare^ 
grunndwall by the side of ffrundy gonndhallf^ but the three last 
may be the result of the triple cons, group. The shortening 
in annd is certainly, and that in sAollde, wollde probably, the 
result of stress-lowering. The ptc sennd shortens by analogy 
of the inf. The following are full examples from O. : 

rl : eorly cherL Back-shortening : harrUg. 

rj?: e{o)rJ)e. Back-shortening: forrperr^ mirrprenn. Excep- 
tions : wurrpy norrp etc. 

rn: bartienn, e{o)meun, lemenn^ stime adj, ^^o)rft«M*, cam. 
Exceptions : berriie * bam,' turmenn, porm, 

rd : Airde, swerd^ birde ( = 0E gebyrd\ Aord, word, bord. Back- 
shortening : girrdelL Exceptions : harrd, gerrde. 

Id : aid, Aaldenn, mlde prt, kald, talde prt, bald, tcilde, milde, 
child, seldenn, weldenn * wield,' feld, geldenn, gildenn vb, gold. 
Back-shortening: allderrmann, chilldre pi, elldre cp, sAulldre. 
Exceptions : sAollde, wollde. 

ng : lang,pwang, sang sb, amang, ping, singenn, springenn, gung, 
sungenn, sprungenn, tunge, king. Back-shortening : anngrenn, ern^ 
glissA, Aunngerr, Exception : brinngenn, 

nd : Aan(n)d, land, f and. prt, bandess pi, mnd, findcnn, bindenn, 
blind, ende, wen{n)denn, blendenn, Aund, mnd adj, nmnde, toundenn, 
fundenn, grand, minde, kinde sb. Back-shortening: hanndlenn, 
unnderr, Aunndredd, vmnnderr. Exceptions : annd, anndsioare, 
Aan(n)d, stanndenn, winndeclut, winndwenn, senndenn, sennd ptc, 
wen(n)denn, gonndAaJlf. 

mb : lamb, wambe, cafnb, climbenn, dumb. Back-shortening : 
timmbredd ptc. 

636. There are other combinations of voiced conss. with a 
preceding vowellike (r and I), which never lengthen : 

rj : birrgenn vb * bury.' 



MIDDLE ENGLISH: VOWELS. 171 

rf: herffesH. 

rm: arrm^ berrme^ louTTm^farrme. 

Ijh : foUghenn^ swolgAenn * swallow.' 

If: Aallfepl 

Im : allmes*. 

687. The eME lengthenings described above are confinned 
for IME by Ch spellings such as queeme = OE cweom, hoord ; 
odd; hoond sb, bounden; doumby the back -shortenings by 
under etc. 

638. The Ch spellings yong^ songe ptc, tonge etc show that 
the older lengthenings had been shortened again before ng. 
This was, perhaps, the result of (r)g) having been reduced to 
simple (i)) finally, as in MnE. 

689. The lengthening of final conss. in strest syllables 
{6%i) was sometimes carried— in the form of doubling— 
into the infected or derived forms, especially in the case of 
words which very frequently occurred in the uninflected 
form. In Ch God has gen. Ooddes, and siij) has pi sAippes". 
wAal^ whale* inflects wialles in North, and NWMl. 

640. Doubling of medial m is very general even in eME 
in summe pi, which occurs in Lay., AR, GE, Wicl. etc. This 
doubling does not appear to be due to the uninflected sum, 
for it appears also in utnummen (AB), comme (KS) = OE cuman. 
Doubling of » is general even in the earliest ME in tinnape 

* scarcely '=0E uneafie, which may, however, be the result of 
some attempt to associate the unfamiliar root-word with ned 

* need.' Other examples are tonnnunge (AR) wonned (Ch)=OE 
wunode. 

For the shortening of double conss. in unstrest syllables 
see 616. 

VOWELS. 

641. In treating of the ME sounds in detail, it will be 
convenient to use diacritics to supplement the defective dis- 
tinctions of the mss. These are (~) and the marks added in 
f and p, with which the reader is already familiar, § being, 
however, used in the value of OE €b=>\,S^ ^= group-lengthened 
and new-long e and Oyd^o with the value of (u), ii z=z u with 



172 HISTORY OF ENGLISH SOUNDS. 

the value (y). The best type of fully developed eME is 
afforded by AB, whose language preserves many archaic 
features lost in O. The following is the vowel-system of AR, 
the corresponding sounds of O. being added in parentheses in 
cases of difference ; those diphthongs which are not developed 
in O. being marked * : 

a, §(a), 9(a) i 

«i (ai) 
au 



e, eo 


u, u (1) 





ei 




oi 
*ou 


e, eo 
ei 


u, a (i) 

ALT', • 

*ui 


6, 9 (a) 


eu 




"'^^ou, *9U 



a 9 1 

9U 

a> $» 9- 

642. The most marked of the ME vowel-changes — as far 
as the strest vowels are concerned — is the smoothing of the 
old diphthongs ea^ eo, ea, eo, a change which — with isolated 
exceptions — was fuUy carried out already in OTr in all the 
dialects except Kt, which preserves the old diphthongs 
throughout the ME period. These ME smoothings keep the 
quantity of their originals, ea was smoothed into short a^ OE 
ea and is being thus levelled under the latter sound, which 
in AR is written e : i^, ^ter, p^t = OE i^, csfter, pat ; ^t^ 
sch§rp = OE eart, scearp. In the earliest Sth texts ea, which 
in AR is restricted to the long sound, is written also — ^inter- 
changing with e — for the short § : heap^feader, feier = OE hap^ 
fader. In Ld a, <?, ea, a are written almost at random : wan, 
v)€8, weaSy was, hafde, hefde, heafde, Aafde, So also in Lay. eo 
for ea in weorp (Ld) prt, weorp (J\jl,=tppp AR) seems to be due 
to the influence of the to. a for ^ is rare in eSth, except that 
in AR it is regular after tr, as in water, hwat, ward = earlier 
we{a)ter, Awet, weard, OE water, hwat^ weard, a in other words, 
such as hlak (AR), may have been taken from the OE inflected 
forms hlacu etc. It is also possible that a was sometimes a 
Fr way of writing the broad x- § survives into IME in Kt ; 
thus Ay. has wes, hedde, vet = OE was, htrfde,fat. Elsewhere 
f became a even before the end of the eME period. In O. the 



MIDDLE ENGLISH: VOWELS. I 73 

change is carried out completely: wasa, haffde^fatt^ old te sur- 
viving only in group-lengthenings (670). 

648. Old a is kept in eSth in such words as varen, Aavefiy 
makien, Aahhen = OE faran^ hafap^ macian^ habhan, ,The § in 
gl^dien by the side of gladien^ beajiien (Jul.) =: OE gladian^ bapian^ 
is probably due to gl^d^ b§J) = OE glad^ ba}, a is constant 
in eSth and general ME am^ where the a is a weakening of 
eo (442). So also a is a weakening of eo and eo in eSth Aa 
* she/ * them * = OE keo^ hare * their/ ham * them '= OE heora^ heom 
(444). a as a shortening of OE a is eSth and general ME 
in such words as halwen {hallghenn O.), garlei=0'E Adlgian, 
gdrleac (AngL -lee). 

644. In chaff arCy chapman = OE ^ceapfaruy ceapmann, a is a 
shortening of f through ^. OE a is shortened in the same 
way, as in laase, naddre =OE UesBa, nmddre (non-WS neddre), 
for which le^e, neddre also occur (671). 

646. The Southern OE ea before ^combinations is pre- 
served only in Et in the form of ( ja) : iald in ES, healde^ yald^ 
yhyealde in Ay., which also has such forms as aUcy boldeliche, 
agreeing with those of the other dialects, which all have the 
Anglian a : eSth a/, ha^\ sail = OE all, hallf, salll, agreeing 
with their retention of the Angl. a before Id (694). Ld 
fluctuates between ea, a and a : call, all, all, half. Here the 
ea is probably due to literary WS, and the ^e to the usual 
graphic confusion between ea and a. 

646. Old p is kept in eSth, as mJ)onhien, moni, name =: OE 
J>gncian, mgnig, ngma. So also in Ay. in some words : Jxmki, 
but many, name. In Ld the unrounding of g is completely 
carried out : man, mani, fram. So also in O. In Ch g survives 
only in the group-lengthenings Iqng, Ignd etc (694), and mfrom, 
which seems to owe its to the analogy of the prepositions 
of, on or to/rp from Scand./ra; elsewhere Ch has a and the 
new-long a: m4in, many, name. In WMl was kept, as is 
shown by such rhymes as mon : on in Harl. and Aud., although 
Aud. writes also man, and rhymes echame : blam^e. We have seen 
(416) that in OMerc. g before ng became u in unstrest syllables ; 
in WMl this change was carried out in strest syllables as 
well before nh and ng, as shown by such rhymes BjAfionke 



174 HISTORY OF ENGLISH SOUNDS. 

vb : gef % idnke (0£ Aince) in Harl., and the spellings lung 
adv (OE Ignge), sung sb in Aud. 

647. Mis i is the regular continuation of OE i, as in smip^ 
ipriten ptc, and is not subject to new-lengtheni];ig in IME (623). 
It represents also the unrounding of OE y in North, and Ml 
(663). eSth has i = OE y in some words, especially after k 
in king (but eilime = OE cynn)^ kinewurfie^ kimej) ' comes' {cumep 
in AR with the u of the iniSn.) = OE cyng^ cynewierpe^ cyifnepy 
also in drihtin^OEk dry Men. 

648. i from e(o) before front conss. is general ME before kt 
(00), as in Uht * lux/ * brilliant/ • easy/ hrihi^ fihten {feUen 
Jul.)=OE l€{o)ht {leokt), beorht, 'breht {$11), f^o)ktan. Other 
cases are dialectal: tiggen^ wri(c)cA€de * wretched' from OE 
slogan, wrfcca, 

649. There is eMnE evidence of a distinction between close 
and open i — ^probably X, I — dependant on the nature of the 
following cons. (786), and it is very probable that this dis- 
tinction was already developed in IME. But, as in IWS all 
Is were rounded into y, it is evident that all the remaining 
in must have been close Is, and this must have been the state 
of things in eME also — at least in Sth. 

e 

660. ME e corresponds regularly to OE e and ^, as in west^ 
ielpen, eten ; rest^ wenien, mete = OE west^ helpan^ etan ; r^t^ 
w§nian. m^te. That OE e [ was levelled under OE ^ in ME 
is proved by the identical treatment of the new-longs eten^ 
mete in IME and eMnE, in which latter the vowel of eat and 
meat was still kept apart — in pronunciation as well as spelling 
— from the ee (ie) of meet,/eeld (^//'i) = OE gemetan, feld, the e 
of the latter word being a lOE lengthening of earlier e = [ 
(395). This broadening of OE e is also shown by such 
spellings as aten, teal, salf=01E etan, wel, self in Ld. 

651. The new long / (including, probably, Orm's e in ende, 
670) is still ke^ apart from the p of «^=0E sa in some MnE 
dialects, and the two sounds must have been distinct in 
ME as well. If f was x*> ^ i^ OE, then e may have been 



MIDDLE ENGLISH: VOWELS. I 75 

either x* or [♦ ; if ^ had been narrowed to x* iii eM£, then the 
IME S can only have been [♦. Cp the evidence bearing on 

^(665). 

662. e is often the result of group-shortening of OE e^ as 
in mette^g€meUe'\ also of ct and ea^ though here it is repre- 
sented also by a (644): 9lejpte^ clensen, birefi=dlapte (non-WS 
slepie), dSentian, bereafod. 

658. In here^OE hire the lowering of the i seems to bathe 
result of want of stress. 

654. ^=0E y is regular in Et^ as in zenne, dede==OKt aenn, 
dede (478), except in king, evel = OE t(fel is WMl as well as Et. 

eo 

655. In Ld the OE eo had been merged in the open e^ as 
is shown by such spellings as erthe^ iamde by the side of 
geomde, early earl, eo itself being written not only for OE eo, 
but also for OE ^, as in feon =ffnn, eeotte prt. In O. eo is 
written pretty regularly for OE eo, the being, however, often 
omitted, so that such spellings as heore and here, weorrj^enn and 
i€errj>enn, heoffne and heffne interchange constantly. In one 
place O. writes h^re to show that the diphthong is short. 

656. In eSth eo is regularly written, not only in such words 
as heore (also hare 643), eorJ)e, heovene, but also— -in agreement 
with Merc, against 1 WS — in weole, weoreld (also world), eleopede = 
IWS toela, woruld, elipode, the eo being, however, in some words 
confined to the earliest texts: thus AR has speken against 
8peoken in Jul. eo is occasionally written for OE ^ ; thus Jul. 
has unweommet, bicheorren—OEk ungeto^mmed, hec^rran. Jul. also 
has eo for Fr e in feovereles * February.' 

657. In other eSth texts we find remarkable fluctuations. 
In Hom. eo is often written 0, as in boren, horte, solf=rQiEk 
be(o)ran, heorte, 9e{o)lf, such spellings as heovene, hevene also 
occurring. This is also common in ON, where it rhymes 
on old eo or ^, as in vorre : iterre = OKfeorran : deorra, hovene : 
Hevene =01& heofon : sle/n, bore * ursum :' spere = OE be{o)ran : 
9pere. The rhyme siorve : orfe is an exceptional one on OE 0, 
which is perhaps due to some change of pronunciation {^eovf 
for orfX). There is also a rhyme of eo on u : honne : kunne= 
OE heonon : cynn. This rhyme-fluctuation between e and y 



176 HISTORY OF ENGLISH SOUNDS. 

seems to point to the intermediate sound (oe), of which is 
a common OFr symbol. In WMl (Harl.) we find such spell- 
ings as Auere=:0^ Aeora, huerte by the side of herte^ Aeovene. 
FPP has simple u, as in Aure, cAurl, duri. So also AllP in 
urpe, bum=01& beam, the form bu^m indicating probably a 
long vowel. It is evident that in WMl old eo had passed into 
a simple sound resembling H, which again points to (oe). 
Conversely, we find eo written for u in Lay.^, as in leore for 
the lure=0^ lyre *loss* of the older text. 

668. It is probable, therefore, that the change of eo into 
the e of IME (Chaucerian) ertAe^ Aerie was not direct (by 
dropping of the 0), but through (oe), itself the result of con- 
vergence of the twto elements of the old diphthong. This 
(oe) was then gradually unrounded, a process which, to judge 
from the orthography and rhymes, must have begun early in 
Sth. What the precise value of eo is in O., is doubtful. It 
is possible that he regarded it as a half-traditional symbol of 
close e — the sound into which eo would first develop — as 
shown by the aimlogy of eo (681). 

660. In Kt eo is preserved as a diphthong, but in the form 
of (je) : yer^e^ ly .ne, wyefde = OE eorJ)e^ leomian, weofod * altar.' 
Simple e also occurs, as in erplich^ sterven. This (je) seems to 
point to an intermediate (joe). 

660. One result of the OE variation between toeo- and 
WO' (4^6) was that some words beginning with original «?(?- 
changed it to weo-. Thus Lay. has weord, weolkne by the 
side of word^ wolkne ; hence our welkin. The shortened wo- in 
wodnes-dai underwent the same change, giving our present 
Wednesday. 

u 

661. is the regular equivalent of OE u, as in sune, and is 
not subject to new-lengthening. In some words, such as ns 
(also ous=us) it ia B, shortening of OE i?. The analogy of i 
makes it probable that IME distinguished between narrow i 
and wide i. 

662. u in some words is a backing of «, the change being 
clearly shown by the spellings with 0, as in mdcAe, *&?^=eSth 
muche{l), swiicA. These spellings appear already in Lay.' 



MIDDLE ENGLISH: VOWELS. 1 77 

This u was, of course, first developed in Sih. (and WMl), but 
it afterwards spread to the other dialects, even Scotch having 
u in muckle, 

u 
668. OE y was completely unrounded in WMl even in the 
OTr period, as is proved by such spellings as birien by the 
side of byrien, sinner, dide=01& hyrga% ^ynna, dyde in Ld. O. 
has ummbe=OE ymbe, but the u is probably due to the Scand. 
«i». Except in such cases of analogy O. has only i : tinne, dide. 
664. In Sth the old y was preserved unchanged under the 
disguise of u^ not only in such words as wnne (which never 
rhymes on tunne 'sol'), diide^ but also in specifically IWS 
forms such as mucAel, htoiichy sullen^ iehiippen=-Wf^ mycel, 
hwylcy syllan^ icyppan^ eWS mieel^ Aioelc, sfllan, tcieppan ; 9uggen 
= OE 9^gan also occurs. U was also kept in WMl, as in 
eii99e : blisse (Harl.), bUgge (Harl., PP1.)=0E bycgan. Even Aud. 
still writes gulte ^O'E *gyltig. u survives as a vaiiant of i 
(and Kt e) in some Ch forms, such as bUrieUy biUy, whose (y)- 
sound is confirmed by eME evidence, together with the pre- 
sent pronunciation (beri, bizi), which shows that the u in these 
words cannot have been made into u. 

o 
666. answers to OE (?, as in on, folk, cole, bodien^OTSt on,fole^ 
col{u), bodian. New-long S in cSle, bddien is still distinguished 
from ^=:0E a in some of our dialects, and is kept apart from 
old in standard E. both in spelling and pronunciation : cp 
coal (koul) with cool (kuwl)=0E col. eSth o must also have 
been distinct from the p j of mpn etc, for this sound was after- 
wards unrounded into a. eME o must, therefore, have had a 
sound between the OE } and the broad j — nl }, which is the 
present Gm sound of short o in ttoci etc. In IME the three 
sounds S, p and o were, therefore, probably }♦, p and }4 resp. 
The MnE dialects seem to point to l« as a later sound of ^. 

Long Vowels : a. 

666. OE a was rounded into g in Sth and Et. The earliest 

texts, such as Jul., still write a : hwa, laverd, gast, but the form 

wumme^wd (u) mi I presupposes the rounded vowel, and makes 

it probable that the a is a traditional spelling, and in so by the 

N 



178 HISTOEY OP ENGLISH SOUNDS. 

Bide of 9wa the is fully established. AB writes and oa^ as 
in hwoy hwoa^ mo(a)re. oa also appears in Lay. {ihoaien\ GE {J,oar) 
and elsewhere. 

667. Hence a occurs only in Fr words in eSth, as in dame. 
In IME a appears as a new-lengthening of OE a, as in hdre^ 
ndme^ mdiien=etl[E Aare, name, makien, OE hara^ nama, maeian. 

668. In Ld and O. OE a is preserved ; tlius O. has Awd, 
laferrd, gdsU In North, a has been preserved unrounded up to 
the present day in the Scotch dialects, where it has been 
levelled under new-long d\ North, texts have only in a few 
cases of apparent borrowing from Sth, as in 90 ^ lord by the side 
of 9wa^ laferd in PC. The later Ml texts show a remarkable 
fluctuation between the North, a and the Sth g. Thus the 
EMI GE rhymes woa: Eva, moal {=mdl): natural on the one 
hand and ggn : on (cp on : don in the same text) sp : temptacid on 
the other. Even the Yorkshire TM rhymes ha/me :fdme (Fr) 
and mgre : befSre. The NWMl AllP rhyme mare : I fare, mgre : 
echSre. In the MnE dialects the rounded vowel has prevailed. 
It is probable that the freedom of rhyme considered above was 
the result of the old a being at first only slightly rounded. 

9 
660. In IME no distinction is made in writing between ^ 

and 9, both being written indiscriminately e, ee. In O. they 

are distinguished as a and e with perfect regularity. In the 

earliest Sth ea is written pretty regularly for the open sound, 

but even in AH e is often written for ea. Great irregularity 

prevails in Ld. 

ME f is the regular representative of the common OE a 
in all the dialects, and of OE ^ in all the dialects except Kt 
(679) ; thus to OE see, l^eran, heafod, hleapan correspond in O. 
e€B, iarenn, hjrfedd, Uepenn, in eSth %ea, learen, heaved, leapen. 
Ld writes meast, meet, behead, bebad=0^ mast, bebead etc. 

670. Orm's a also appears as a lengthening of OE a (ea), 
as in am, ard^O^ am * house ', card * country '. 

In dame * secret' it is a lengthening of OE g {d§me, IWS 
dyme), which in E. words such as ende, sendenn is written with 
simple e. In gerrsalam, Elysabap it seems also to be a lengthen- 
ing of OE^ — a lengthening which may have begun in OE itself. 



MIDDLE ENGLISH: VOWELS. 1 79 

671. The shortening of common OE ^ appears in O. some- 
times as a, sometimes as e : 

a : wrappCy lasae^ lasslenn, laffdlg, 

e : fie9%h (also flrssh\ clenfisenUj ledde prt, spredd ptc. 
The regular shortening of OE ^ in O. is evidently a. The e of 
clennsenn and ledde is really a shortening of e (cp O.'s clene, ledenn^ 
§ 676), and the same may be the case with spredde^ although the 
r would tend to preserve the ^ (674). f£S8h^ lastly, may owe its 
e to the following (originally) front cons. (733). The other 
EMI and North, texts have generally e, lesse, for instance, 
occurs in rhyme in North., GE, and TM. The earliest North, 
also has lefdi against the a of O. The e of eSth and Et is 
ambiguous, but the evidence of the later texts — in which it 
becomes a — shows that it stands for f. The ON rhymes 
foranne : monne^ AR has wrastlen, RGl has amti^ laddre, WMl 
also has a : clad in AUP, lasse in rhyme in Aud. Ch. generally 
has a, as in ladi, ladde, wrastlen. Sometimes he shows the 
EMI ^, as in lesse^ lasse^ both forms occurring in rhyme. 

672. represents the common OE e in all the dialects, as in 
her, medsy iene=0^ her, med, cene (cmie). The Anglian <?=WS 
^ and le (IWS y) appears also in all the dialects to the ex- 
clusion of the WS forms, except that ^=1WS occurs in some 
of the extreme Western dialects (690) ; thus O. and Sth agree 
in such forms as sellg, sell, efenn, even, dede^nonW^ geselig, efetiy 
did, WS gesalig, of en, dad; sene, kere^in, heren=^ nonWS gesene^ 
geheran, WS gesiene, gehleran (IWS gesyne, gehyran). Here, again, 
Ld shows great irregularity : let, leot, ket prt, to geamene, atywede 
= 0E Ut, to gemenne (WS le, y), afetcde (WS le, y), the last 
spelling being evidently a WS literary one. 

ME e in feld, sheld is an OE (Mercian) lengthening of e. 
O. also has e as a lengthening of OE ^ (against dame 670) in* 
be{o)ide7in 'encourage', weordenn * injure* = OMerc. geb^ldan^ 
dic^rdan (IWS gehyldan, dwyrdan), where tf^=^ (681). 

673. In the following words O. has <:5 = Gmc and WS a, ea in 
parentheses indicating a confirmatory eSth spelling : 

(a) before r: har {GB,),j)ar (ea), icarenn prt, hwter,f{sr,gar, bare 
* bier.' The only exception is that gar is sometimes written ^^r. 

N 2 



l8o HISTORY OP ENGLISH SOUNDS. 

(b) after p: strate, rad (ea), radenn, dradetm, redenn, dredenn 
also occur, gredig always has e^ being the only complete ex- 
ception. The rule is further confirmed by Sth breap and MnE 
thread. 

{c) before 1 : mal, selig has only e. 

(d) after 1: keche^ Imtenn^ blatenn^ skepenn. The subst ^laip 
has also e. No other exception. 

(e) after w : w^epenn. 

674. It is evident that in these words the a is due not 
to any WS influence but to the low tone of the vowellikes 
r, /, w — influences which had already been partially developed 
in OE (449). In spache by the side of specke and dadbote by the 
side of the uncompounded dede there is, however, no such 
influence, and these forms may really be WSaxonisms. It is 
remarkable that these e^s are more developed in the Mercian 
O. than in the Saxon Sth dialect. 

676. The shortenings of Angl. ^= WS m show exclusively e in 
Ml and North.: O.hss ermdey redd ^tcdredd-pto; bleddre^deppte ; 
neddre. Ch has a in naddre^ bladdre^ and has both dred and drad 
in rhyme. AR has neddre^ bleddre, which probably means \ — 
the forerunner of Ch's a. It is uncertain whether the e of blest 
(Ay.) was long or short. GE and TM have blast in rhyme ; 
this exception to the general rule may be due to the' influence 
of the vb blawen, 

676. In a few words O. has, on the other hand, e for common 
OE <§, especially before n : lenenn, menenn, dene (but clauTiesse). 
iseness and inusne have a. The other cases are : del (but dalenn), 
lefedd ptc, ledenn. It will be observed that, except in the case 
of lefeddy all these ea are followed by a point cons. 

677. ME e in all the dialects (with some exceptions in Lay. 
and Kt) also corresponds to the Angl. er-smoothings of O WS and 
OKt ea, eo (46a, 465). Thus O. has ec, Aek, nek ; le^henn^fleghenn 
=OAngl. ec, heh, neh; Ugan^flegan (WS eac, heah, neah; leogan, 

fieogan), Orm's eo in seoe by the side of sec^J)^o)8 * thighs' may 
be a merely orthographic variation (681), aided by some 
tradition of WS orthography. The earliest Sth shows the 
same forms as O. : eke, heh,fiehe sb (AR^). For the later de- 
velopments heih, hlh etc see 696. O. has exceptionally a 



MIDDLE ENGLISH: VOWELS. iSl 

= WS ea before back conss. in some verbal forms — dak^Jlah — 
where they may be due to the analogy of such preterites as 
bad=OTSt bead, where the a is regular. 

678. In Kt the lOKt ^=0E y is preserved, as in loeV (KS), 
ver {Aj.)=zOE Awyjyr. 

670. The diphthongic ea is preserved in Kt in the same 
spellings as the short ea : great, diad, dyad, lyeave (also grot, belave 
etc), probably with the value (jaa), although if the second 
element had been lengthened we should have expected ''^^^o^ etc. 

680. In 9cawen (Ld)=OE sceawian the e seems to have 
been absorbed by the preceding front cons. (733), the length 
being shifted on to the a, our present 9how pointing to 
ichdwen, although AR's scAaioen points to short a. O. has d=. 
ea in drah^OEt dreag * suffered' and lafe =0E geleafa 'belief*. 
Are these modifications of ea or of later a ? 

681. In Laud the old eo had evidently been completely 
merged in i, as shown by such spellings as cesen, (/er=OE 
ceasan, dear, and, with the usual confusion between close and 
open vowels, dor, eo being also written for OE e, as in leot prt. 
The treatment of eo in the other texts is also quite parallel to 
that of eo. O. writes preost, prest, deofless, defless, AR writes 
preo9t, deovel=OE preost, deofol. O. also occasionally writes 
eo for ^, as in dreofedd=OE gedrefed {gedrqfed), Galileo =i Galilee. 
So also AR in cheoken * cheeks '=OMerc. cecan (WS eeacan). 

682. The spellings oe and are common in Hom. and ON. 
The latter has Jhede : noede= OEJ)eode : nede, ko = heo * she '. Harl. 
has seo, se *see' : me, deor, duer, hue, he *she', /«rtf=OE hleor 
' cheek '. PPP has duj), buf = OE deoj), beoj). AllP have bot 
:=beodeJ) ; their ho * she' probably =^55 from ^hjo (685). These 
spellings point to a convergent smoothing into {♦ in Sth, h in 
WMl, the latter being preserved in the MnE chme (choose is a 
very late half phonetic spelling) from OE ceosan. AR^ writes 
eo for the Fr (oece) inpreoven, where AR^ has pruvieu. 

688. In Kt «> is represented by the diphthongic (jee), 
probably from earlier (joeoe), as in chiese, chyese, byefi pi, for 
which byfi etc also occurs. This diphthong also represents 
OKt S in some words : ihierde * hired ' =a OKt geherde (nonKt 
gehyrde) in KS, kyer adv, ih{y)erd * heard' in Ay. 



1 82 HISTOEY 0¥ ENGLISH SOUNDS. 

684. In O.'s frnvwer, Sth vour=^OE feower the e seems to 
have been merged into the o by the influence of the two lip 
conss. 

686. 0.'s^>?J=OE hed 'she' is the result of stress-shifting 
in weak syllables (44:^), the stages being fi[*>, 2c}t, £(d}*, o}*. So 
also in the word %he itself (733), and in you^ your. OE eow, eower 
became (joou, joouer), whence, by the usual di-opping of the 
initial glide, the eSth ou, duer, 0.*s gure keeps the first half 
of the old diphthong, and changes the second half into u by ' 
the analogy of the poss. pm of the ist pers. (ure), O. then 
makes ^goro (or *gdw) ixiioguw by the analogy ot^ure. 

I 

686. answers both to common OE *, as in tcls^five, and to the 
Angl. group-lengthened i in such words as childy slngen^fhiden^ 
climben, t before ng being shortened again in Ch, as in syngen. 

For the i in hlk etc see 696. 

a 

687. answers to OE u, as in huSy mup^ Ch hom^ mouthy and to 
the Angl. group-lengthened u in such words as 9u?tgen, Au?id, 
dumb, the H being shortened again before ng in Ch : fdngen, 
houndy doumb. 

For the u in fuel see 696 ; for that in gure see 685 ; and for 
the IME u in enough see 721. 

ii 

688. The old y was completely unrounded in EMI, as shown 
by such spellings B&for-pi,forr'J)i,fr inLd and 0. = OE. foT-py^fyr. 

680. In Sth and WMl it was preserved in the Fr spelling 
tt(i), as in Auiren * hire' (AR), iujfen * make known* (AR),/«r 
(AR, Harl.),/«*^ (A11P)=0E kyran, cypan,fyr,fy8L 

600. In WMl ^ also represents IWS ^= Angl. #. Thus PPP 
has huren * hear', nudful * needful ', which appear in AR etc in 
the Angl. forms hereuy neodful. 

For a from eo see 68a. 

601. The main source of ^ in ME generaUy is the OFr u and 
uiy as in ciire, fortune^ due, friit, fruity the latter spelling being 
the most usual. U final or before a vowel became eu {eu ?) in 



MIDDLE ENGLISH: DIPHTHONGS. 1 83 

Ch, as shown by such spellings as vertewy crewel^vertu^ cru§L 
So also OFr iu in eschue^ eschewe ' eschew '. 

5 
602. answers both to common OE J, as in do, mone, god, and 
to group-lengthened Angl. 0, as in word, gold. 

608. In North, o was fronted to {♦, Fr £♦ being levelled 
under the new sound, as shown by such rhymes as 9Wie: 
foribne = fortune. 

For the in two etc see 695. 

9 

604. is the regular representative of OE a in Sth, Et and 
in later Ml (668), as in mgre, Apm, as also of group-lengthened 
Angl. a in such words as l^ig, ipnd, cgmb, 

606. g after w became o in IME in most words, as in twd, 
who, womb, as shown by the MnE pronunciation (831). wgd 
' woad ' is an exception. 

Diphthongs 

606. We have seen that the OE diphthongs disappeared in 
all the ME dialects except Et, but their loss was supplied by 
new developments. All the common ME diphthongs are the 
result of various changes in the combination vowel + the A>1- 
lowing OE conss. : w (/), h (c and o), open g e and g ©. The 
combination vowel + w and of back vowel + back h and g yields 
a diphthong of the (au)-typ6, the combination front vowel 
+ front h or p yields a diphthong of the (ai)-type. Thus OE 
diaw, dohtoTy dragan; heh (Angl.), we§ appear in fully developed 
ME as d^, douiter, drauen ; heih {hih), wei. It will be seen ihat 
there are two ways in which these diphthongs are developed : 
(i) by weakening of the cons, into a glide- vowel, as in drauen ; 
(%) by parasiting, as in heih, where the glide from the [« to 
the o has developed into a full glide-voweL The second 
process is generally the most primitive one, and it is some- 
times doubtful — as in the case of wei — whether the second 
element of the diphthong is not really a parasite-vowel which 
has absorbed the original cons, rather than a weakening of 
this cons. The last stage in the development of the ME diph- 
thongs is the absorption of the glide- vowel into the preceding 



184 HISTORY OF ENGLISH SOUNDS. 

vowel — an absorption which is inevitable in the ME weak- 
ening of OE ug etc, as in fuel from OE fiigd through futod. 
Sometimes this absorption is the result of the assimilative 
influence of the glide- vowel itself, as in hth from heihy where the 
i first drew up the preceding ^ to t, and was then absorbed by it. 
607. As the combination vowel + ^ is in itself scarcely dis- 
tinguishable &om a diphthong of the (au)-type, and as the 
combination front vowel + ^ had become almost — if not quit© 
— a diphthong even in OE, it was natural to keep to and g as 
symbols of the second elements of diphthongs. This is done 
in O., which, at the same time, shows the development of the 
diphthongs in its most primitive stage. In O. i and back 
(open) g — ^which he writes gh — do not develope parasites: 
dohhterti heh^ draghenn. The second elements of the (au) and 
(ai) diphthongs are expressed by w and g resp,, which are 
doubled after short vowels, not only finally and before conss. 
but also between vowels: clatowesi (=0E elawa\ daw; wegg, 
leggd ptc (=0E jel^gd). In these doublings between vowels 
the first cons, denotes the glide on to the vowel, the second 
that vowel itself (^clauwess). The fact, however, that O. did not 
adopt the latter spelling shows the doubling was really a kind 
of phonetic fiction to enable him to mark the shortness of the 
preceding voweL That clatowess meant practically nothing but 
clauea is further confirmed by O.'s spellings aicw for Lt au^ and 
egg for Scand. ei^ as in Aww8tin=Au{gu)stin^ hegglenn * salute' 
= heila. In the ptcc ilagenn, (forr)legenn=iO^ slagen, {far)legen 
the^ is left undoubled. Conversely, it is sometimes doubled 
after a long OE vowel, as in tweggenn=twegen. So also the to 
is doubled in cAetvfoenn=OE ceowan against ne{o)we=^0¥t (Angl.) 
neowe etc. It is doubtful whether these doublings indicate real 
shortening of the preceding vowel, or are merely the result of 
confounding length of vowel with length of glide on to the 
second element ; it seems, on the whole, most probable that these 
(as hiAofowwerry ovmhar^OE^ feower^ ohwar) are cases of back- 
shortening (629). The doubling of ^ after i, as in drigge^^OE 
dryge is merely a way of marking the length of the vowel : 
drigge:=>dTlgey or rather drle, That^ had been completely ab- 
sorbed by a preceding i or I is made probable by the occurrence 



MIDDLE ENGLISH: DIPHTHONGS. 185 

of «^e=OE Age * victory* at the end of the line (610) and the 
speHing n^rfasit^ as also by the insertion of ^ in such forms 
as dHgcrafft * sorcery' =0E drjcraft, ZacaHge by the side of 
ZaearCe, Hence ig in -ig -lig^ probably represented simple i : 
ia/i^=(haalii). 

698. In the other eME texts — as also in later ME generally 
— the second elements of the diphthongs are represented by i 
and «, as in Lt and Fr. Already in Ld the Fr begins to prevail 
over the OE spelling : dages^ dai; foiaer^fau, treuthe. In eME 
retention of the cons, symbols often seems to indicate that the 
diphthong is not fully developed, but the revival of «? in the 
IME combinations aw etc shows that the system of spelling 
carried out in the O. was never completely disused. In North, 
the consonantal spellings egh etc, as in degAe=deie * die', were 
kept up and revived, in order to avoid the ambiguity of «i, ei 
etc, where the i in North, had come to be a mere mark of length. 

600. Diphthongs are occasionally formed by the develop- 
ment of a parasite-i before various front conss. besides A and ^. 
Thus AR has leinteUy acweinte=zOE l^ten, actc^ncte. 9c (733) 
has the same effect in aiscAe, waucAen (Wicl.)=OE asce, wasean. 

§i(ai) 

700. is the regular eSth representative of OE ag. Thus dag 
appears in Laud as dag, daig, dai, and mag, lag, appear in Lay. 
as mat, leai, lei, AR has generally ei : dei, mei, lei, teide. 

701. In O. the first element has undergone its regular change 
into a: dagg, magg, lagg. So also in slagenn=islein (AR). This 
at is still rare in AR (dai), but it occurs in Lay. (saide), being 
frequent in Lay.' (^n^y)) and it occurs even in Ld {daies gen.). 
The eSth ai is probably due to acoustic divergence rather than 
to the isolative change of ^ into a. This is confirmed by the 
fact that Kt, which otherwise preserves §, agrees with Ch and 
IME generally in having day etc. 

702. Scand. ei (and gy) becomes ai finally, except mjiei, as 
in Orm's nagg, «w^=slcel. nei, mey * maid'. Lay. has nai, AR 
and North, nai, Ch nay. This change is against that of non- 
final Scand. ei into egg, ei (705). Was Scand. ei pronounced ir 
when final ? 



1 86 HI8T0BI OF ENGLISH SOUNDS. 

708. The eSth eii from OE ah, as in seiA, eihte from Angl. 
gesah^ ahta seems to have been ei rather than ^, for these 
words generally keep their ei in MSSt—^eigh (sig)^ eighie^ although 
say also occurs (in rhyme in Ch). 

For the at of aische etc see 699, 733. 

ei 

704. is the regular development of OE eg, ^g : thus OE weg, 
gelegen, regen, §ge ' fear ', l^gde appear in O. as we^, legenn, ^^^«, 
egge, le^de^ in AR as wei, ileien, rein, eie, leide. Orm*s se^de 
(seide in AR) from OE sagde has taken its e from the pres. 
slogan. Lay. has seaide pointing to OE ag, 

706. ei also represents unfinal Scand. ei, as in Orm's Peggre, 
hegglenn, beggtenn=Ice\, J)eira^ Aeila, beifa, final Scand. ei ex- 
ceptionally inj)egg * they*, perhaps because it was unstrest. 

706. In IME there is a tendency to confuse ei with ai, es- 
pecially in North., where the bldest mss write wai, thai, thmr 
etc. In Ch the distinction between day and wey is still kept 
up, but there is a tendency to confuse them, ey being ofbener 
levelled under ay than vice-versa, thus we find alwey rhyming 
on fey (Fr), pley and alway rhyming on day, abhay (Fr). ai had 
probably begun to front its first element into \, which would 
bring the two diphthongs very close together. 

au 

707. It is not improbable that dw etc were diphthongs 
already in OE in such forms as sdwle (also written saule), and 
hence also in ME. Otherwise au does not appear in the earliest 
ME except in foreign words, such as Avmstln (O.), sawter (AR) 
* psalter '. 

708. OE. dragan appears in O. as draghenn, in the earliest Sth 
as drahen, in AR and Ch as drawen, the e having been first 
rounded into 6, which by a slight relaxation of the back of 
the tongue becomes iq w. In drawen the to was probably soon 
weakened into an u. The back h was rounded in the same 
way in ME, and developed a parasite n before it in AR, where 
drawen has pres. drauifi=OTw!s dragAefip. So also Orm*s 
laiAgietin — where the^A is perhaps meant to indicate the round- 
ing of the Ak (cp ^aAge in Lay.^) — appears in AR as lauhwen. 



MIDDLE ENGLISH: DIPHTHONGS. 1 87 

- 709. IME au etc 81*6 sometimes the result of a change of v 
into a lip-open cons, and then into 10, as in kani=eilLE Aavei, OE 
ic^oc. mauk 'worm' from eME ma/&e^=Icel. majjk seems to 
have passed through the stage of *mavek^ and then to have 
followed havek, 

710. The correspondence of ME ai and au to OE ag and ag 
respectively has sometimes given rise to doublets. Thus we 
have slagenn, slein^ slayn from OE slagen on the one hand, slawen^ 
slawe from OE slagen on the other, some texts, such as Ch and 
RBC showing both forms in rhyme. 

on 

711. The development of om in ME is quite parallel to that 
of au, OE boga appears in Lay. as ho^e^ ^J^L ^ ^^ *s ^^^^ 5 ^ 
to Orm's dohhferr^ brohhte correspond AB*s douhter^ hrquhte. In 
Ch the u is often omitted in boghfe etc to prevent confusion 
with o«=(uu), being implied by the following gh. OE Um 
retains its spelling unaltered to the present day. 

712. ou in eME is the regular representative of Scand. gu : 
thus O. has rotowst^lcel. f-aust 'voice', the earliest Sth has 
lawmen * loosen ', formed from the verb IpyM^ but with the vowel 
of the adj. Igui. 

eu 

713. The only regular source of this diphthong would be 
OE eu>, as in str^w'tan, but it is rare. Exceptional IME ew fr*om 
OE ^in ewte * newt' from efete. 

fi 

714. OEfage * fated' appears in Lay. ZAf^Bige^faie^feie^faie, 
Other examples are ei *egg' (AR), keie (Ld)=OE ag^ cage. 
In all of these words the a is Angl. as well as WS. 

ei 

715. Angl. eg appears in O. as 4?, egk^ which also represent 
Scand. ggj, eig (through Dan. a;, eg\ iareg(A)enn, eghcy leghenn^ 
degenUy leghe ' hire '= Angl. wregan (older wr€sgan\ ege (WS eage\ 
legan (WS leogan)^ Icel. de}/ja^ leiga. The earliest Sth shows 
the same forms in many cases, but often also with the change 



1 88 HISTOfiY OF ENGLISH SOUNDS. 

of e into *, of which there are already traces in OAngl. (465), 
thus kB} has eke^ liken. KR? has toreien^ eie, li{g)eny dei{g)en. 

eou 

Tie. answers to OE «w, as in eSth heou^ neowe =Ajigl. Aeow, 
neawe (WS M10, niwe). O. has hewe, neowe^ newe. In IME this 
diphthong becomes eu by the regular change of eo into e ; thus 
Ch has Aewe, newe, 

Ol 

717. druie in AR=OE dry^e must have had this diphthong 
once, although druie may be equivalent simply to drue (595)« 

fu 

718. answers to OE aw, Saw as in Orm's lawedd, daw^ sAawenn 

m 

= OE lawedf deaw^ sceawian. Jul. has le(a)wede, AB has 
^cAea(u)wen and scAawen (680). 

eu 

719. See eou (716). Angl. ^=WS aw would give eu in 
ME, but the combination occurs very rarely ; bilewen (Horn.) 
' betray ' is an example. 

ou 

720. was first developed out of OE ow^ as in 9idw * place', 
fiowan^ which appear in ME as stowe^flowenn (O.). Such forms 
as inouAy drouA, touward, nouAware = OE genoAy drog^ tdweardy 
noAwar are fully developed in AE, but not in the earlier texts, 
which have only inoA etc, as in 0. In the last two words the 
u is afterwards dropped. 

721. In Ch ou in the combination ouA becomes (uu) : j/nougA, 
slougA =(inuux, sluux), as shown by the MnE forms (897). 

For (?w=OE eow see 685* 

9u 

722. is the regular development of OE dw^ dg, and of OE a 
before A, Thus AR has cnowen, owen, auA * ought ' = OE cndwan, 
dgen, dA, O. has cnawenn^ agAenn etc. In North, this diphthong 
does not round its first element, but remains au : knauy awen. 



MIDDLE ENGLISH: CONSONANTS. 1 89 

just as OE stdn remains stan, Et and WMl have the same 
diphthong: zaule, knawe in Ay., crawe Harl. in rhyme, cnatoe in 
AUP. In these dialects the want of rounding is probably the 
result of shortening. 

CONSONANTS. 

728. The following is the consonant-system of fully de- 
veloped Sth : 

TBBOAT BACK FBONT FORKW. LIP 

h h h - )?, s, sch f, wh 

k ch t - p 



g 

n(g) 


Z 

(1) 

g 


r 
1 
d 
n 


f , S V, ^ 

b 
- m 


c 





— 


u, 8, t\ >, 



QO 



e 


(T) 


(0 


V,8 


>, 


— 


(«) 


00 


- 


- 


Q 


mcD 


m 


— 


B 


d 


(1-) 


1 


— 


r 



724. The OE dropping of unstrest A (500) led to its complete 
loss in the case of the pronoun AU in Ml and North. While 
AR, Kt and Ch preserve Ait {Ayt)^ O. writes iU, North, it, this 
form occurring already in Ld. In these dialects the rare 
emphatic Ait was supplanted by the very frequent unemphatic 
it in writing as well as speech, against the analogy of Ae, Aim, 
whose frequently-occurring emphatic forms were made the 



I90 HISTORY OF ENGLISH SOUNDS. 

graphic Bymbols of the weak and strong forms alike. An 
interesting instance of the loss of weak h is afforded by the 
eSth ending -Ud in fosirild * fosteress ', generally used in a 
depreciatory sense, as in mapelild ' chatterer '. It can only be 
explained by the OE name^ in -iild, which survived mainly 
in the poetry, and would naturally suggest such parodies as 
mapelild of MSepkild etc. 

726. Of OE ir, hly hw, An only Aw was universally kept in 
ME. The old spelling Aw is kept in eSth, becoming Au in Et, 
but O. reverses the elements, writing wAillc for the AwUcA of 
AB and the AuicA of Ay. This shows that OE Aw must have 
already assumed its present sound of o. In North, the back 
element was exaggerated, giving a, a pronunciation which 
was indicated by writing qu — quilc. This spelling is also found 
in Ml texts, such as AUP and GE. 

726. lA and «A=OE Al^ An still survive in the Ay.: lAard^ 
nAdfe=0'E Alaford, Anutu. rA, however, has become simple r 
in Ay. : reg=.0^ Arycg, O. has occasionally such spellings as 
Ihude^ rAdf, but generally writes simple r, I, Ld drops the A 
not only in such words as Idverd, wile, but also in wua^Awd^ 
wat. Similar droppings are common in many other early 
texts, as also of A before a vowel. It is probably mere 
carelessness in many cases, due partly to the loose usage of 
Fr scribes. The addition of A before a vowel is not uncommon 
in eSth texts. 

727. Uninitial A was in OE split up into two sounds c and 
o. The former of these was rounded into G in ME (696). 
The front A, which occurred after front vowels, is sometimes 
written s in eME where it occurs before t. Thus Lay.^ has 
driste. The Proverbs of Alfred have drislin^OE dryiten, other 
examples being mistie, ristewis. Here the s is an imperfect 
representation of the high pitch of o. drofte=OE broAie in 
Lay.^ is an attempt to symbolize the rounding of c. Lay.^ 
often writes / for both sounds : an AeJ)^ on high ', cnij)te, Jtorfi 
* through *, brojjte. The cons, is often omitted entirely in these 
early texts, even Lay.^ having such spellings as almiten, broute. 
This can hardly indicate an actual loss of the conss. them* 
selves, but is rather part of the general looseness in the 



MIDDLE ENGLISH: CONSONANTS. I9I 

writing of A, and also of that unwillingnesB to use it in a 
strong oonsonantal value which afterwards lead to the general 
use of ^^. 

f, B, f 

728. That initial/, #,/ were voiced in Sth and Kt is proved 
by the initial v of AR and Ay., and by the initial z of Ay. 
Ay. keeps i before cons., as in ilage, smal^ although he must 
have pronounced z here also, as shown by the MnE dialects. 
Fr/and 9 are not voiced, as hnfol (A'R),J^ite (Ay.), Mvf * safe ' 
(Ay.), which shows that the voicing of the native initial 9 etc 
must have been developed before the nth cent. Words 
which were introduced before the Conquest were naturally 
assimilated to the E. pronunciation^ being so few in number. 
Hence AB. has v in vah. Ay. has z m*za^n Jon (alternating 
with 9) by the analogy of 0£ 9an{c)t with its (z). 

The MnSth dialects have (z) not only initially and medially, 
but also finally, as in (guuz)=OE gds. This shows that the 
final 9 of Ay. has no more value as evidence than his frequent 
medial s in ase, prayie by the side of aze^ prayze. Final z is 
found in AllP : sydeZy gemmezy he ISvez, he sez. Final /* in AR 
was a graphic necessity (602), and proves nothing. We may 
assume that /, #, / were voiced everywhere in Sth and Et, 
except, of course, in such combinations as st. 

720. The present E. voiceless pronunciation of final /, s,f 
must have been developed in ME before the loss of weak e, 
for the distinction between (bsAp) and (baa'Sz, hoi's) can only 
be explained by the ME bafi, bapes^ bap(i)e(n) (baa)>, baa'Saz, 
baa'Sd). Hence, although Ch's final/ in stafno more proves a 
breath sound than his /in of prp — still pronounced (ov) — yet 
we must assume final as well as initial (|>, s, f ) in his mainly 
Midland dialect. When this Ml and North, breath pronuncia- 
tion began — ^whether it began initially or finally, or simul- 
taneously in both positions, whether it was already developed 
in OE, and whether, if so, it existed there (in the Angl. 
dialects) from the beginning — there is no evidence to show. 

730. In MnE we have initial ("S) in weak words, such as 
the, that, then, though. So also finally in with. The prp qf also 



192 HISTORY OP ENGLISH SOUNDS. 

has (v), contrasting with the (f ) of the adverbial (j^— both 
from 0£ cf. We have (z) in originally inflectional syllables, 
as in h(m%e% (hauziz), iree%^ contrasting with the (s) of goo9e^ 
geese. The exceptional voicing in all these cases is evidently 
the result of want of stress. It probably began (or was kept up) 
between vowels and voiced conss., the/ in such collocations as 
tofe . . . , onJ)e.,, being treated as an ordinary medial /. For 
the parallel y= weak ch see 928. 

731. Hence every unstrest weak monosyllable with ("8, z, v) 
must originally have had a corresponding strest or strong 
form with (|7, s, f ). We still preserve this distinction in our 
of and off, and the older pronunciation (wi)?) for (wiiS) is no 
doubt the remains of a similar distinction, which was not 
kept up, because no divergence of meaning or grammatical 
function had developed itself, as in the case of of and off. 
Such rhymes as blisiis in Ch, wace {=was)=face in AllP seem 
to point to a similar distinction between strong (is, his) and 
weak (iz, hiz). 

782. In MnE we do not hesitate to use the original weak 
(hiz) etc as strest emphatic forms also. That this was im- 
possible in ME is shown by the d of quod=0^ cwafi. The 
otherwise anomalous of quqfi (Jul.) as opposed to ctoefi (AB) 
can only be explained as the result of want of stress (cp 418). 
As the word was mainly used as an enclitic, the strong form 
quafi died out in most ISIE dialects. When the weak (kwoS) 
was made emphatic, the anomaly of final (S) in a strest 
syllable was got rid of by the change of (^) into (d). 

sch 

783. That OE sc had become a simple sound different from 
i in ME is clear from the spellings sA, ss. The remarkable 
spelling 9ca 'she ' =0E seo in Ld is the earliest one that points 
to some such pronunciation as our present sL The a is 
merely an inaccurate spelling of e, of which there are many 
examples in Ld, and the development of the form must have 
been something like (sjoo, sjoeoe, sjee, Jee) with the same 
change of (sj) into (J) as in the MnE 9ure (915). The develop- 



MIDDLE ENGLISH : OONSONAKTS. 1 93 

ment of the 0£ 9c in 9cori^ Mcan pi etc must have been 
similar : 8Q, so, s\ t, dear evidence of the froiit stage being 
afforded by the parasitic i in aiscAe=0^ ascan. 

784. The Scand. sk before front vowels no doubt had a 
fronted if but this fronting must have been very slight, for 
the Scand. sk is generally preserved in ME before all vowels, 
as in skin, ihl, iii, except in a few words of early introduction, 
which followed the analogy of the OE sc, such as shiften. 

785. In North, the unstoest -M of the ending -Uh becomes -m 
in /»^/f>=: OE ^nglisc. So also North. «a/, 9uld=. OE scealy seolde 
appear to have been originally weak forms of the emphatic 

ii 

786. In Stii and Et there is a tendency to drop all -^eak 
final ns, not only in inflections (especially verbal), as in lAnde 
inf. ibunde ptc, but also in derivative syllables, as in game, game 
= 0E gamen. North., on the other hand, keeps all its final /rs, 
thus showing exactly opposite tendencies to what it did in the 
OE period (53a)- 

ch 

787. ME ehz^OE 6 i£(, when doubled, wtitten cch, cheh, such 
spellingB as etretche (Wid.), fetche (TM) occurring only in 
isolated instances in IME. This seems to show thai OE 6 
had not-^in eME at least-developed into full (tj). Probably 
it had the sound of qo, which is that of Sw k before front 
vowels, as in hnd * cheek.' 

788. Initial eh occurs before the following OAngL vowels 
{Si^i examples marked f being from 0. : 

80: ^chaff, ^chegsire^caf iyf^ ceefj^ daeter (WS deaeter). 
i : ehireke, ehihen^dirice, eieen. 
e: ekeite^deit (WS cieet). 

j=WS ie (469): eherren *tum,* ^ehele^dfrran, 6 fie (WS 
dierran, ciele). 

ea : eharhin ' grate ' = ceareian, 
eo: "fekerl^^dearL 

•=:Gmc J: ek^essceee (WS eieee), 





194 HISTOBT OF ENGLISH SOUNDS. 

e=WSltf: eiepm * aeSl* ^dspan. 
i: eAiden^ feAild^acuiany 6Ud. 

ea {e) : "fcAappmann, cASisr^ciapmann, dice (WS ciace). 
eo (i) : fcAe&enn^eeosan. 
730. Initial i remains before conss., as in clap^ cuiht^ and 
before the foil. AngL vowels : 

a, 9 : ^eare^ ^callf^ \cann = caru^ calf ( WS cealf)^ cgnn. 

u : feumenn = cuffuin. 

o : eole = coL 

a : t>^U = eald (WS C(?a&;). 

H : eu =z cU. 

0: cdl = col. 

f = WS §1 -f'iemfnpe ^champion' = cpnpa, 

y: "fiinff = cyning^ cyng^ iicAene = cycene. 

y : fiipenn ^ make known ' = cyfian, 
cb : ftine = c^^ (cene). 

740. Traces of the non-Angl. fronting before eal + cons, are 
seen in the ^t cAald, cAold = WS ceald, and in cAali by the 
side of cali = WS dedlc. The i of kerven &om ceoffan may be 
due to the infl. of the ptc corf en. The c^ of Orm's ptc cAosenn 
= OE coren, is, on the contrary, due to the prs and prt cAesentiy 
cAa9 = OE eiosoHf cioi, whence also the #. The A of ietel — 
Prompt, has both cAetil and ketUr^bxA of Orm's Aifrke can only 
be explained by the infl. of Scand. k§tU^ kirkja, whose is were 
only slightly fronted, and were therefore levelled under 
HEk. 

741. In Sth and Et non-initial cA, ccA correspond to OE c^ 
cc preceded by mutated vowels, as in mUcAel^ micAel, wreccAe = 
OE micel (Qoth. mikU)^ wrpcca (from *torakkfo). In O. and 
North, we often find k answering to the Sth-Kt cA. Thus O. 
has wreccAcy but mikell. The only exceptions in Sth-Et are the 
result of OE c being immediately followed by a cons, which 
hinders the development of the front hiss, thus in AR tfcAen 
has 3. prs sg t^kp^ and in Ay. zecAen has 3. prs sg zekp. The ex- 
ceptions in 0. are partly explainable by analogy, or by the 
infl. of Scand. forms. Thus wirrkenn against Sth taircAen may 
owe its i( to the sbst toerrky and mikell may owe its i to the 



MIDDLE ENGLISH : CONSONANTS. 1 95 

Scand. miiil. But neither explanations apply to such a word 
M^ennkenn, with its 3. prs sg J/enie/fp and ^ri Jiohhie ; no Scand. 
word is close enough in form and meaning to influence it. 
The correspondence of ekenn^ bisennienn with OAngl. eian eto 
(535) seems, indeed, to show that the absence of fronting is 
older than the period of Scand. influence. It is possible that 
the regular development of Orm's dialect was to change all 
non-initial es into k, and that the cis that occur are due to Sth 
influence. It is worthy of note that three words which have 
cA have also the special WS a =s Angl. e : lacAe, spiecAe, wracAe. 
Ch shows his usual compromise between Ml and Sth in his 
distribution of cA and A. Thus he has both iiien and secAen^ 
beteien and besecAen in rhyme, and recien and reccAen. Mn£ 
generally prefers the informs — seek, reck — beseecA being an ex* 
ceptional Southemism. 

742. The development of cA after front vowels is a difficult 
question. The comparison oictoik = OE cwie with the Mil dia* 
lectal quitcA ' couch-grass '=0£ ctcice (from *kmkd) shows that 
this influence requires to be helped by a following front vowel 
— which, if Qmc, would make the cA fall under the previous head. 
That a following back vowel stops the fronting is shown by 
such eSth forms as Igdlukest^ Igdluker^OT^ Idplicott^ Idplicor con- 
trasted with IgdlicA, adv lpdlicAe=OE lafilic, lafRce. The final 
cA of 4icAy as also of the sbst HcA ' body ' = OE Rc^ and oipicA 
= OE jpic may be explained from the infl. of the inflected 
forms 'lice etc. Orm*s fluctuation between Re and llcA^ bacc 
and baccA = OE bac points to an OE gradation tie^ Itces^ bac^ 
bade. But this will not explain the Sth icA. Here the cA 
seems to be the result of want of stress, which would enable 
the preceding front vowel to carry out its influence without the 
help of another front vowel. This may also be the explana- 
tion not only of the cA of -licA^ but also of that of the Sth stoicA, 
AuncA, ^cA, ilcA =: OE ficilCf w?elcy AwUc^ Awdc^ die from ^iwalik 
etc. Also of the wicA = shortened OE icic in GreenmcA etc, and 
perhaps of ditcA by the side of dyke from OE die. The form 
Aic in KS — Ay. has icA — ^and ice in O. may have been origin- 
ally the strong form corre8i>onding to the weak icA. But O. 
has k also in 4ike = Sth -RcAe. 

o % 



196 HISTOET OF ENGLISH SOUNDS. 



748. The dropping of c in i — which thus becomes the weak 
form corresponding to what then becomes the strong icA, ice, 
finally superseding these latter — cannot be direct; fmust rather 
be referred to the ONorth iff (540). O. shows a similar weak- 
ening in ^lig; := Sth -licA, OE -/kr, as in gditilg = Sth ggfUichy 
0£ gdstlic ; our present 4]/ can only be referred to Orm's form. 



3 

744. OE g becomes i everywhere in ME, except in the com- 
bination n^ and dg, where Lm, dUD graduaUy developed into 
their present sound of (ns, ds), as in sengen, brigje MnE (sins, 
brids) = OE ^ptgan, hrycg. So also in Sth liggen^ leggen, ^eggen 
=: OE licgan, l^cgan, secgan, which in Sth rhyme on brigge, Fr 
allegge etc. The evidence of the MnE dialects shows that in 
O. and North, these words returned to their original back 
conss. : leggenn^ ieggenn. 

745. g = Gmc j always becomes ^, as in ge, gnng = OE §e, 
geong, iung. 

746. Initial^ = Qmc^ occurs before the following OAngL 
vowels (cp ck), examples marked f being from O. : 

»: t^«^ = ^^(WSytfa/). 

i«' pfi ^ gift' 

e : -fgellpenn ^ boast ' == gelpau (WS gidpan), 

^ = WS iei -f^errde r= g^d. 

ea : fgarrienji * prepare * = gearcian. 

eo : gelwe = geolu, 

e : -fgemenn = geman ( WS gieman). 

I: td^^^rr * greedy ' =z gijre. 

ea (e) : -fgan = ongian. 

eo (e) : ^getenn = gSoian. 

747. Initial stopt g remains before conss., as in grene^ gnag^% 
and before the same vowels which preserve initial ki gaUe, 
gengen * go,' gUt, fgat (pi of gdt * goat '), get * geese ' = OE galle 
(WS gealle), ggngan, gyU, gat, gch. 

748. Of the exceptional initial gs, some are Scand. words, 
such as gerp 'girth' s= Icel. gjprp. gest ^ WS giest is also a 
Scand. form ; cp geest * yeast ' = OE gest. The vb biginnenu in 



MIDDLE ENGLISH: CONSONANTS. 1 97 

O., which has g in eSih also, gets its g from the pit Ugann and 
ptc higunnenn. As higetenn keeps its ^ in 0., it is difficult to 
see why its pret. should be higait {higft AR) with a g against 
go,f (with occ. gaf\ which, again, does not agree with the g of 
gifenn by the side of gifenn. As eSth shows exclusively^ in 
bigeten, given^ it seems possible that the unanalogical g% of O. 
are due to Scand. influence. But on the other hand, there is no 
such verb as ^bigeta in Scand., and Scand. grfa has a diflTerent 
YOweL North, has give^ gette. Ch, as usual, hesitates between 
Norl^em and Southern : yiven^ but geten. The MnE gidd = 
Orm's geldenn^ OMerc. geldan ( WS gieldan) no doubt owes its y — 
against the g of give and get — to the fact that it has lost the 
old strong forms answering to OMerc. gdld^golden, which would 
otherwise have introduced the g into the inf. and prs. O. 
fluctuates in gate^ gate, which may, perhaps, reflect the OE 
alternation in gat (WS geat), pi gatu. eSth has, of course, 
get. Ch has gate. North, has yate^ thus reversing the usual 
relation. 

740. There is a tendency to drop initial g before i, especially 
in weak syllables. Already O. has if by the side of gif. 
Uikel = OE Ugieel is also ME. giccAen has dropt its ^ in MnE 
itcA. In all these instances^ = Qmcj, OE initial ge^ (Qmo 
gi' ga-) becomes f- in ME, as in inoh (O.), ivere AB. = OE genoh^ 
§ef8ra. 0. still has such forms ssgeAdtenn, 

750. Non-initial open g and g are represented in O. hygh 
and g resp., the latter probably representing a vowel in most 
cases (697). g occurs after OE front vowels Anally or before 
a cons., as in da^, ^^9 ^^^1 snd before another OE front 
vowel, as in legenn ptc, wregepp = OE wrosgep, gh occurs after 
and before an OE back vowel, as in inoghe pi, dagken pi, nigkenn 
= OE genoge, dagos, nigon, and after n /, as in burrghess pi, 
follghenn = OE bwrga.folgian, A following OE back vowel or 
preceding cons, changes original g ^ gh^ even when mutation 
has passed through it into the preceding vowel ; thus to wregep} 
corresponds the infin. wregAenn^OE wregan^ and not only 
serrghenn * grieve * = OE scsrgan has gh, but also serrghepp = OE 
icergep, -ip. i = WS ea, eo acts like the diphthongs of which it 
is a smoothing (463, 465), and keeps ^A before all vowels, as in 



1 98 HISTOEY OP ENGLISH SOUNDS. 

egAe, ligAenn^ le^hefifi = OAngl. ege^ legan^ legep (WS 9age^ l9ogan^ 
*leogep). 

751. In Sih and later ME generally Orm's gh after front 
vowels ifl levelled under g ; thus to his egge * fear,' ighe corre- 
spond eie^ eie in AB, the latter becoming ye in Ch. gA after 
back vowels and conss. is written A in the earliest Sth — a 
spelling which occurs also in lOE — as in daAeiyfuAd^folAep^ to 
in AB and later ME generally: dawes^fuwd^ valuwep, pointing 
to the development e , 6, 8. 

752. Final gA becomes A. Thus to Orm's plurals inogAe, 
burrgAe correspond the singulars inoAy burrA. Hence, by a 
natural analogy, original final A^ as in ^A, became gA before a 
vowel — pi Aegke^ spl AsgAe%9t — as already in IWS Aeage^ AeagoH, 
So also OE AdAyfurA became in IME Aolwe^furwe. 

768. OE g after r and / preserves its front character not only 
in Sth, but also in the later Ml and North., being vowelized 
to i, as in burien^ birien = OE byrgan. Ld also has bebiriend. 
O. itself has birgenn^ but as it occurs only once, it is probably 
a scribal error for birrgAenn. The i of birien is probably a 
parasite- vowel, which was already developed in OE bebyr(i)gan. 
The i which regularly represents final g after a cons, in 
ME, as in milrit meriy belt = OE myrg^ hK^)9» ^ ^^ doubt this 
parasite. 

t, d 

754. Weak final d is regularly unvoiced in earliest Sth. 
Thus Jul. has inempnet^ naket^ totoart = OE genpaned^ nacod, 
Umeard. The later Sth texts restore the d^ thus KB? has 
offered against the offearet of AB^ = OE cffared. It is probable 
that the d was preserved in earliest Stii also before a vowel 
beginning the next word, the change into t taking place only 
before a breath cons, or a pause. This unvoicing of weak stops 
— ^which may be of OE origin (cp 533) — ^is fixed in the MnE 
contracted participle3 dwelt^ wU etc. 



MODERN ENGLISH: PERIODS. 1 99 



MODERN ENGLISH SOUNDS. 

PERIODS. 

765. It is still more difficult to draw a definite line between 
late Middle and early Modem E. than between 0£ and eME. 
The most marked criterion is, no doubt, the loss of final e in 
Mmey ndm&» etc. The loss of final e — of which we see the 
beginnings in Ch, and which was completely carried out by 
the middle of the 15th cent. — broke down the metrical system 
brought to perfection by Chaucer, and made a new departure 
necessary. The break between old and new was made more 
abrupt by the social confusion caused by the Wars of the 
Boses (1450-71), which, at the same time, helped to level 
differences of dialect — at least, in the upper classes. When 
printing was introduced — ^in 1476 — the language had almost 
completely settled down into its Modem, as distinguished 
from its Middle, stage. The difiusion of printed books made 
the want of a common literary language more and more felt, 
and, at the same time, greatly facilitated the realization of the 
ideal — an ideal which was, however, not fully realized till the 
appearance of Tindal's translation of the New Testament in 
1525 — a work which is wholly modem both in vocabulary 
an<f diction. 

756. We may, then, say that Modem English begins, in 
round numbers, about 1500, the period between 1400 (or 
rather later) and 1500 being regarded as Middle Transition. 
The change from ME to MnE is, with the exception of the 
loss of final <?, slight compared with the changes in MnE itself. 
Even if we separate the language of the period from 1800 to 
the present day as 'Living English,' we still require a division 
of MnE into tiiree periods, which may be conveniently desig- 
nated as First, Second, Third, Living English itself requiring 
a twofold division : 



200 HI8T0BT OF ENGLISH SOUNDS. 

' 1500-1600 First Modern English (fMn) 
1600-1700 Second Modem English (sMn) 
1 700-1 800 Third Modem English (thMn) 
180P-1850 Earl^ Living English (eLE) 
1 850-1 900 Late Living English (ILE). 

767. These minute divisions are necessarily even more 
arbitrary than those into OE, ME and MnE. The separation 
into centuries is mainly for the sake of convenience: in 
reality fMn extends some way into the following century, and 
if MnE were to be separated into two periods only^— Early 
and Late (eMn, lMn)-ri65o would, perhaps, be the best point 
of division, agreeing with the general upheaval caused by the 
Civil War. thMn is really a transition to LE, because its 
sounds are still more or less known to us by tradition. It is 
also to be noted that our knowledge of LE really extends 
some decades beyond the present time, because the observsr 
tion of the tendencies of vulgar speech enables us to predict 
with some certainty the future deyelopment of the standard, 
educated speech. 

768. The E. of Tindal and his successors was not a mere 
literary language— it was a spoken language, which every 
educated man acquired more or less perfectly, whatever his 
native dialect might be. Even in the 14th cent, we find the 
Kentish man Gower writing — and probably speaking — a dia- 
lect which, in spite of some marked Kenticisms, is practically 
that of the Londoner Chaucer. In the i6th cent we find 
natives of Wales, Lincolnshire, Cambridge, London describing 
the sounds of one and the same dialect, although, of coifrse, 
the influence of the native speech shows itself occasionally, 
as it does still, in, for example, the pronunciation of an 
educated Yorkshireman. We have, then, in MnE to recognize 
H standard E. (stE) as distinguished from dialectal E. 

769. The question now arises, where was this stE developed ? 
The answer is easy. Ch was a Londoner; and his dialect 
was such a compromise between EMI and SthE as would 
naturally be spoken in the capital — at the court, and by 
the educated classes generally. Chaucer's disciple, Occleve, 
was also a Londoner. The succeeding poets, Lydgate, EEawes, 



MODEBN ENGLISH : PEBIODS. 201 

Skelton, were all EMI men, the two first being natives of 
Suffolk, the last of Norfolk. This movement towards the 
East and North is clearly shown in the language as well as 
the literature. We may, therefore, define stE as that mainly 
EMI dialect of ME which was developed among the educated 
classes in London, and thence spread to the Universities, and, 
in more or less dialectally modified forms, over the country 
generally. The influence of stE in Scotland was purely 
literary. Although this influenee was strong enough to make 
stE the liturgical language of the country, it did not extend 
to speech, for even in the last cent, pure ' Broad Scotch ' — 
which is really Modem Northumbrian — ^was the conversational 
l^^guftgd of educated Edinburgh, and even now educated 
Scotch has a sound-system which is wholly distinct from that 
of stE. The educated speech of Ireland has also a sound - 
system of its own, which is an independent development of 
eMn, influenced by the Celtic Irish. The educated speech of 
America is analogous to that of Ireland, being, like it, in some 
cases more archaic than the stE of England. The educated 
speech of Australia and New Zealand is only beginning to 
diverge from that of England. 

760. In all our large towns there is a marked divergence 
between the speech of the upper and lower classes, which is 
most marked in London. This difference between stE and 
vulgar E. (vgE) extends over the whole English-speaking 
world, many vulgarisms of London E. reappearing not only in 
the popular speech of Birmingham and Liverpool, but also in 
that of America, although, of course, each town has its own 
vulgarisms. Vulgarisms are of various kinds. Some of them 
are due to the influence of neighbouring dialects. Others are 
archaisms, which once formed part of the standard language ; 
and others, again, are anticipations of changes that are 
imminent in the standard language. Hence the necessity 
of the study of vgE (by which is here understood the vgE 
of London) both as preserving the fossilized standard pro^ 
nunciations of an earlier period and as pointing the direction 
of future changes. 



202 HISTORY OJ" ENGLISH SOUNDS. 



PHONETIC AUTHORITIES. 

701. The orthography of MnE is a direct continuation of that 
of ME. ME orthography itself was, as we have seen, highly 
unphonetic in its basis. In MnE the divergence between 
sound and symbol increased. Thus abready in fMn e had 
not only the ME values (e, ee, e&),but also that of (ii). But the 
application of this unphonetic basis was still mainly phonetic. 
The influence of tradition became, however, stronger and 
stronger as the printing-press developed, until the printers 
became the main arbiters in questions of orthography, their 
interest being, of course, to make it as uniform and con- 
servative as possible. By degrees, not only the basis, but 
also the application of E. spelling became unphonetic. Al- 
ready in fMn final e was written at random, or used as a 
mark of length of the preceding vowel, and by the end of 
sMn there were so many silent letters (such as gh)^ and so 
many isolated correspondencies of sound and symbol that 
elaborate spelling-rules became necessary. Meanwhile the 
orthography became more and more fixed, settling down in 
the beginning of the next period into practically its present 
form. 

762. But, whatever its present condition may be, MnE or- 
thography was never intentionally unphonetic in its period of 
development. On the contrary, a number of spelling-reformers 
arose in the i6th cent., whose avowed object was to regulate 
and simplify E. spelling by restoring the direct connection 
between sound and symbol. The new alphabets proposed 
were, however, without an exception, too intricate and cum* 
brous for practical use, which, indeed, is not to be wondered 
at, when we consider what difficulties these reformers had to 
face, and how utterly unprepared they were to grapple with 
phonetic and alphabetic problems. But, although they were 
not able to provide a workable substitute for the unphonetic 
French basis, they succeeded in introducing some important 
improvements of details, such as the separation of u and t;, ea 
and ee — all of them purely phonetic reforms. Although most 



MODERN ENQLISH : AUTHORITIES. 2O3 

of the reformers were men of high education — including in 
their ranks such classical scholars as Cheke — they were not 
much troubled with etymological considerations. K they 
tolerated the silent 9 in island^ it was simply because Fr 
orthography had familiarized them with the use of # as a 
mark of vowel-length, its introduction into this particular E. 
word being, of course, directly suggested by the identity of 
its meaning with that of the Fr isle. 

768. Hence even the ordinary eMn spelling has a distinct 
value as evidence of changes of pronunciation, and often 
serves to confirm and control the statements of the phonetic 
authorities, and their phonetic transcriptions. 

704. Although eMn spelling is still some guide to the 
history of the sounds, it is quite inadequate by itself: our 
main reliance must be on the phonetic treatises, which, 
fortunately, become more and more accurate and reliable as 
the fixity of the spelling leaves us in the lurch. Some of the 
sMn authorities, indeed, show an acuteness and accuracy of 
analysis and description of sound- forination which partly 
anticipates the discoveries of Mr. BelL The statements of 
the f Mn authorities on the formation of sounds are, on the 
other hand, mostly vague and confusing ; and here we have 
to rely mainly on their comparisons of English with foreign 
sounds — mainly French. Unfortunately, Fr pronunciation 
itself has changed even more than E., and the statements of 
the older French orthoepists are as vague as those of their 
English contemporaries. It is, therefore, fortunate that we 
have detailed comparisons of the sounds of f Mn with those of 
a phonetically written language whose sounds have under- 
gone hardly any change since the i6th century — North 
Welsh. The results thus obtained are further confirmed and 
supplemented by a phonetic transliteration in Welsh ortho- 
graphy of a Hymn to the Virgin^, the mss of this Welsh 
transliteration (HVg) having apparently been written about 
1500. 

1 Phfl. Soe. TnuiM. 1880-1, *35. 



204 HISTOBT OF ENGnSH SOUNDS. 

765. The following is a list of the phonetic authorities from 
the 1 6th centiuy downwards in chronological order ^. 

First Modern Period. 

1530. Palsgrave, John (Pg). 

LeBclarcissement de la Langae FrancojBe. London. 

This book is in E., though the title is id Fr. Fg graduated 
at Cambridge, 0:xford, and Paris. 
To the French reprint is added a reprint of 

An Introdnctorie for to leme to rede, to pronounce and to 
speke French trewlj etc. 

By Giles du Guez or du Wes, with no author's name, except 
as shown by an initial acrostic, and no date, but apparently 
about I53JZ. 

1547. Salesbury, W. (Sb). 

A Dictionaiy in Englyihe and WeUhe . . . wherevnto is pre- 
fixed a litle treatjfe of the englyihe pronunciation of the let- 
ters. London, 

Sb was bom in Denbighshire, studied at Oxford, and settled 
in London. 

1555. Cheke, Sir John (Ck). 

Joannis Cheki Angli de pronunciatione Oraecae potiseimiun 

linguae disputationes cum Stephano Yuintonienai Episcopo. 

Basle. 

The Oospel according to Saint Matthew . . . translated from 

the Greek, with original notes, by Sir John Cheke, knight etc 

... by James Goodwin. London, 1843. 

The spelling in the latter is not strictly phonetic, but rather 
an attempt to improve the existing spelling. 

1567. Salesbury, W. 

A playne and familiar Litroduction, teaching how to pronounce 
the letters in the Br^rtishe tongue, now commonly called Welsh 
.... London. 

1568. Smith, Sir Thomas (Sm). 

De recta et emendata lingvsa anglicas scriptione, dialogus. Paris • 

^ Ellii, Early Bngliah FrommoiAtion, Part I. 



MODBBN ENGLISH : AUTHOBITIES. 205 

1569. .Hart, John (Ht). 

An Orthographie, conteyning the dae order and reason, howe 
to write or painte thimage of mannes yoice, most like to the 
life or nature. Composed hj J. H. Chester, Heralt. London. 
1573. Baret, John. 

An Alvearie or Triple Dictionarie, in Englishe, Latin and 
French. London. 
1580. Ballokar, William (Bll). 

BuUokan Booke at large for the Amendment of Orthographie 
for English speech. 
1605. Erondell, Peter V 

The French Garden : for English Ladyes and Gtentlewomen 
to ^alke in. Or, A Somn^er dayes labour. Being an instmc- 
tion for the attayning vnto the knowledge of the French tongue. 
London. 
1609. Holyband, Claudius** 

The French Littelton. A most easid, perfect and ahsolvte way 
to leame the French tongue, Set foorth by CiavcUve Hotyhand^ 
Gkntil-homme Bourbonnois. London. 
161X. Cotgrave, Randle. 

A Dictionarie of the French and English Tongues. London. 
161 1. Florio, John. 

Queen Anna's New World of Words, or Dictionarie of the 
Italian and English tongues, collected, and newly much aug- 
mented by J. F. 
1 6 19 first ed., 162 1 second ed. Gill, Aleitander (G.). 

Logonomia Anglica. Qui gentis setmo faciliiis addiscitur Con- 
scripta ab Alezandro Gil, FaulinA Schols magistro primario. 
Secundb edits, paulb correctior, sed ad vsum communem ac- 
commodatior. 
1633. Bntlar^ Charles (Bt). 

The English Grtonmar, or the Listitution of Letters, Syllables, 
and Words in the English tongue. Whereonto is atmezed an 
Lidez of Words Like and Unlike. Oxford. 
X640. Jonaon, Ben. 

The English Grammar. Hade by Ben Johnson. For the 
benefit of all Strangers, out of his observation of the English 
Language now spoken, and in use. 

Jonson 'Was bom in 1574. 

^ Ellis, p. aa6, note. ' BUii, p. 127, note. 



206 HISTORY OP ENGLISH SOUNDS. 

Second Modem Period* 

1 663- Wallig, John (W.). 

Joannia Wallisii Grammatioa Lingruae Anglicanae Cvi prae- 
figitor De Loqvela; sive de sonoram omniym loquelarivm 
formatione: Tractatvs Grammatico-PhysicvB. Editio Sexta. 
London, 1765. First ed. 1653. 

1668. Wilkins, John (Wk). 

An Essay towards a Real Character, And a Philosophical 
Language. 

1668. Price, Owen (P.). 

English Orthographie or The Art cf right speUingy reading, 
pronouncing and writing all sorts cf English words. Oxford. 

The author's name is given on the authority of the British 
Museum copy in which it is pencilled. 

1669. Holder, William, D.D., F.B.S. 

Elements of Speech, an Essay of Inquiry into the natural pro- 
duction of Letters with an appendix concerning persons Deaf 
and Dumb. 

1685. Cooper, C, A.M. (Cp). 

Grammatica Linguae Anglicane. London. 

1688. Miege, Quy, gent. (Mg). 

The Qreat French Dictionary. In Two parts. London. 

1 70 1. Jones, John, M.D. (Jn). 

Practical Phonography : or, the New Art of Rightly Speling 
(sic) and Writing Words by the Sound thereof. And of Rightly 
Sounding and Reading Words by the Sight thereof. Applied 
to The English Tongue. 

Third Modem Period. 

1 704. Expert Orthographist (EO). 

The Expert Orthographist : Teaching To Write True English 
Exactly, By Rule, and not by Rote. According to the Doctrine 
of Sounds. And By such Plain Orthographical Tables, As 
Condescend to the Meanest Capacity. The Like not Extant 
before. For the Use of such Writing and Charity Schools 
which have not the Benefit of the Latin Tongue. By a School- 
master, of above Thirty Years Standing, in London. Persons 



MODEBN ENGLISH: AUTHORITIES. 20^ 

of Quality may be attended at their Habitations; Boarding 
Schools may be taught at convenient times. London : Printed 
for, and Sold by the Author, at his ^ouse at the BlueSpihea in 
8fread^EagU-C(Airt in OrayB-Inn^La'M. Where it is also 
Carefully Taught. 

17x0. Falatines. 

A Short & easy Way for the Falatines to learn English. Oder 
eine kurze Anleitung zur englischen Sprache zum Nutz der 
armen Pfalzer, nebst angehangten Englischen und Teutschen 
ABC. London. 

1 7 10. Dyche, Thomas. 

Quide to the English Tongue. London. 

1725. Iiediard, Thomas (Ld)^ 

Grammatica Anglicana Critica, oder Versuch zu einer voUkom- 
menen Grammatic dor Englischen Sprache, in welcher . . . eine 
neue Methode, die so schwer gehaltene Pronunciation in kurtzer 
Zeit zu erlangen, angezeigt . . . wird . . . durch Thoma» Ledtard, 
N.C.P. ft Philol. Cult. Hamburg, 1725. 

1766. Buchanan, James (Bch). 

Essay towards establishing a standard for an elegant and uni- 
form pronunciation of the English Language, throughout the 
British Dominions. London. 

The author was a Scotchman, and there are Scotticisms in 
his pronunciation. 

1768. Franklin, Benjamin (Fk). 

A Scheme for a New Alphabet and reformed mode of Spelling. 
Complete Works ... of the late B. F. London, 1806. vol. n. 

The pronunciation here given is, of course, affected by 
American provincialisms. 

1780. Sheridan, Thomas (Sh). 

A General Dictionary of the English Language, One main 
Object of which, is, to establish a plain and permanent Standard 
of Pronunciation. 

The author was an Ldshman, but familiar with the standard 
pronunciation. 

i Ellii, p. 1040. 



\ 



208 HISTORY OF ENGLISH SOUNDS. 

OETHOGRAPHT* 

766. The two main sound-changes in the transition from 
ME to MnE are (i) the dropping of unstrest e in endings, and 
(2) the shortening of double medial consonants. These changes 
had already been carried out in eNorth., where the unstrest e 
does not cotmt at all in verse, and where we fiild such spellings 
as calls (CM)=OE cealliafi^ bigines (MH). Generally, however, 
both in eNorth. atid eMn the consonant-doubling was kept as 
a sign of the shortness of the preceding voweL This naturally 
led to doubling conss. which were ori^jially written single in 
ME, when preceded by a short vowel. This is rare in North., 
where the doublings in such words as littel^ goddeSy commyng^ 
wonnyng (CM) correspond to real doubling in Chaucer's dialect, 
although all these words had single conss. in OE as well as in 
North, itself. But in such MnE spellings as penny ^ sorrow the 
doubling of the conss. was never anything but a sign of vowel- 
quantity. The dropping of final e in such words as faUe inf., 
lesse led also to the doubling of final as well as medial codss. 
to show that the preceding vowel was short, not only in fall 
and lesSy but also in smally glass etc = ME snud^ glas, 

767. At first there was great confusion in the writing of the 
e and the doubling of conss. The e was often written after 
short as well as long vowels, as in hyme by the side of hym^ 
though in such cases its significance was generally neutralized 
by doubling the cons., as in sonne:=^OEi sunu. e was always 
kept after (v), because this cons, was generally written u^ as 
in ME, through the greater part of the eMn period. We still 
medianically retain this usage, writing final e in have (haev) 
as well as in lehave (biheiv), love (lav) etc. In eMn such a 
spelling as *tou Would have suggested our law (lou). In our 
present spelling we use « as a lengthener only when a single 
cons, precedes, but in eMn such spellings as cAylde=ME eitld 
are tiot unfrequent. The following examples frotn Tindal 
will give an idea of the irregularities of f Mn usage : 

fare (=ME fare), care . life (=^) . cAylde {^chUd) . iooke 
(=:tdi). 

hyme, hym (=^m) . live {= liven), love (z=:luve). 



J 



MODBEN ENGLISH : OBTHOGBAPHY. 2O9 

sionei, sirets {=-sfgn€9y streteij . ax {=axe). 

euppe, cup (=ciippe) .penny, peny [^peni\ hoddy^ body (= 
lodi) . openned (= opened). 

all, ledd ptc, gospell . worship, toorsAippe . aun, sunne (=zsun?ie), 
Bonne {=9une). 

clooeke (=clgie), goodda {—godeij. 

neety nettea (=net, nettes), beed (=bed). 

768. In MnE ci becomee the regular doubling of i. aah {JU- 
%her in Td) is simplified to %h. /is sometimes doubled initially 
(to indicate the breath sound ?), as vo^ffor (Td), a usage which 
still survives in some surnames. 

769. The irregularity in the use of silent e and of cons.- 
doubling in eMn was, as we are expressly told by Sb, kept up 
for the convenience of the printers ' in consideration for iusti- 
fiyng of the lynes.' 

770. The ME use oly for % was carried to a great excess in 
eMn, the two letters being used almost at random, except that 
i was rarely written finally, such spellings as thi for thy being 
exceptional. Final i was also written ie, not only in such 
words as /2V=ME lie, but also in manie, -lie etc. The present 
use of the Fr c to denote (s) in E. words was begun in f Mn, 
the older % being also kept; thus Td has on9, once, tkry%e, 
pence, f alee. For the MnE ea, oa see 817, 831. 

VOWELS. 

771. The changes from ME to MnE are so gradual, that in- 
stead of starting from a f Mn vowel-scheme, it will be more 
practical to take each IME vowel separately, and trace it down 
toLE. 

a 

772. Sb says of the Welsh a that * it hath the true pronun- 
ciation of a in Latin,' and that it is never sounded ^so fully in 
the mouth as the Germaynes sound it in this word wagen* 
Again he says : 

< A in English is of the same sonnd as a in Welsh, as is evident in 
ihese words of English jllb, aal, oeryisia, palb, po/ol, salb, «aZ.' 

P 



; 



2IO HISTORY OP ENGLISH SOUNDS. 

Here the last should be saal ; Sb in his phonetic transcriptions 
often, but not always, doubles the vowel to show it is long, 
and doubles the following cons, when the vowel is short. 
These three examples are all of long a, but in other places he 
gives us transcriptions of short a, thus narrowe : narrw^ %parrowe 
: sparw, kwarter^ hand^ fflaca. The present sound of Welsh a is 
3, ||f , which is also the standard North Qm sound. In Saxony, 
however, a has the deep sound of j (sometimes j ?), which, of 
course, is the one alluded to by Sb. To judge, indeed, from 
Lediard's (1725) identification of the North Gm a with the K 
a in fall this j was formerly universal in Gm. It is, therefore, 
clear that Sb pronounced £. a as 3« HVg has the same tran- 
scription as Sb. 

778. Pg (1530) says : 

* The Boundyng of a, whiche is most generally vsed through out 
the frenche tonge, is suche as we vse with vs, where the best 
eDglysshe is spoken, whiche is lyke as the Italians sounde a, or they 
with VB, that pronounce the latine tonge aryght.' 

Here, again, the ItaUan a is pure ]. Fr a is now ]i- and j, but 
the 1 6th cent. Fr grammarians state that it was clearer than 
the o-like German a, Pgs ' correct ' pronunciation of a was, 
therefore, the same as Sb*s. But he tacitly admits that there 
was another pronunciation. What this pronunciation was, 
we seem to learn from his contemporary du Guez : ' Ye shal 
pronounce your [French] a as wyde open mouthed as ye can, 
your e, as ye do in latyn, almost as brode as ye pronounce your 
a in englysshe.' This points to a sound between j or ] on 
the one side and i on the other, that is, to 3^ or, more probably, 
to the X of our man, a sound which, as we shall see, was fully 
established in the next cent. Equally clear is the statement 
of Erondell in 1605 : 

' Our A is not sounded . . after the rate of the english word ale, for 
if a Frenchman should write it according to the English sound, hee 
would write it in this wise esl, and sound it as if there were no s* 

774. The question now arises, May not Pg and Sb have had 
the same x-sound, and identified it wrongly with the 3 of other 
languages ? This would be possible with Pg, but hardly with 



MODERN ENGLISH : VOWELS. 2 1 1 

an accurate observer like Sb, who was perfectly familiar with 
both of the languages whose sounds he compares. On the 
whole, it seems safest to assume that fMn a had been fronted 
— certainly as far as ]>-, and probably as far as x — in the London 
dialect, but that the tradition of the older ] was still kept up 
by the influx of provincial speakers, bo that the two sounds 
really existed side by side. It is to be noted that, according 
to Butler (1633), short a, as in man^ Aal, had a different sound 
from long a, as in mane, hate. Does this point to 3*-) X^ ? T^ 
Danish the short and long a diverge as ], 3>^) which would be 
precisely parallel. In Swedish, however, it is exactly the 
contrary : 3, J* (almost ji). But see 780. 

775. Wallis (1653) distinguishes nine E. vowels, three gut- 
tural (j, X, 1)> three palatal (x, [, [♦), and three labial (}♦, 1, £♦). 
Of the palatal vowels he says: 

' YocaleB Palatinae in Palate formantur, a^re scilicet inter palati et 
linguae medium moderate compresso : dum nempe concavum palati, 
elevate linguae medio, minuB redditur, qu^m in gutturalibuB proferen- 
dis. Suntque in triplici gradu, prout concavum magis minusve con- 
trahitm:. Quae quidem diversitaB duobus modis fieri potest; vel 
fauces contrahendo, manente lingua in eodem situ ; vel faucibus in 
eodem situ manentibus, linguae medium altius et ad interiores 
palati partes elevando : utrovis enim mode fiat, vel etiam si utroque, 
perinde est. 

'Majori aperhir^ formatur Anglorum a, hoc est d exile. Quale 
auditur in vocibuB, hat^ vespertilio; hate, discordia; jpoZ, palla Epi- 
scopalis ; jpofe, pallidus ; Sam (Samuelis contractio) ; same^ idem ; 
la/mb, agnus ; 2afn€, claudus ; da/m, mater (brutorum) ; daiM, domina ; 
bar, vectis ; hare, nudus ; 6an, exsecror ; hane^ pemicies ; etc. Dififert 
hie BonuB a Gkrmanorum d pingui seu aperto ; eo quod Angli linguae 
medium elevent, adeoque aerem in Falato comprimant ; Qermani vero 
linguae medium deprimant, adeoque a^rem comprimant in gutture. 
Galli fere Bonum ilium proferunt ubi « praecedit literam m vel n, in 
eadem syllaba ut enUndemerUf etc. Cambro-Britanni, boo Bono solent 
Buum a pronunciare.' 

In another place he says : 

'A plerumque pronunciatur sono magis exili quam apud alias 
plcrasque gentes : eodem fere modo quo Gallorum e sequente n in voce 

P a 



212 HISTOET OP ENGLISH SOUNDS. 

entendement, sed paulo acntius et clarius ; seu ut a Italornm. Non an- 
tem ut Germanorum d pingue ; quein sonum nos plerumque exprimere 
BolemuB per au vel aw, si producatur ; aut per d breve si corripiatur/ 

776. This description of an open vowel formed by the middle 
of the tongue and palate points distinctly to our present x 
in man. The clear back vowels ], j were evidently unfamiliar 
to W., who only knew the extremes j and x, and hence con- 
sidered the Welsh and Italian 3 as a variety of x* Ai^d identified 
the Qm j with j. But he does not actually confuse x ^^^ % 
for he expressly says that the E. a is thinner in sound than 
the foreign a, Wilkins's description is vague, but not incon- 
sistent with W.'s. 

777. Cooper's (1685) list of exact pairs of long and short 
vowel sounds is as follows : 



I 
(a) ean 
(6) cast 


2 
ken 
cane 


3 

weal 


4 
foUy 

faU 


5 
fall 

foale 


6 
up 


7 
meet 

need 


8 
foot 
fool 


= I 


c 
I* 


X 


. J 


1 


1 


I 


i 



%B, was possibly X) and 2b may have been [«. 4b is now j«, 
and may have been narrow even in Cp's time. 6a may have 
been 1. 7a and 8a were probably half-long rather than strictly 
short. It will be observed that Cp was dissatisfied with the 
traditional pairing of f £«, i i«, and imagined that I [«, i }4 
were the true pairs, not being familiar either with the true 
short narrows £ 1 or the long wides b h. This identification 
of f [, h [♦ resp. is still a common error both of theoretical and 
practical phoneticians (174)- Cp says : 

* A formator k medio linguee ad concavum palati paululdm elevato. 
In his can poesum, pass hy praetereo, a con-ipitur ; in caM jacio \ past 
pro parsed prseteritas, prodncitur. FrequentissimnB auditur hie sonns 
apud AngloSy qui semper hoc modo pronunciant a latinum; ut in 
amaham. Sic etiam apud Cambro-briiannos ; quandoque apud GaUos ; 
ut in ammal, demomde^ rarb autem aut nunquam apud Germanos. 
Hunc Bonum correptum & productum semper scribimus per a; at 
buic character! prseterea adhibentur sonus unus et alter : prior qui 

* Printed yae«o. 



MODERN ENGLISH: VOWELS. 



213 



pro vocali ejus longa habetur, ut in eane, definitur sect, sequent! ; pos- 
terior ut in UHM sect, septlnia sub gutturalem.' 

' E formatur k lingua magis elevati et expansft qu4m in a propriiis 
ad extremitatem, unde concavum palati minus redditur ft sonus magis 
acutus; ut in ken video. Sic apud Germanos menschen homines. 
Apud GaUoa rarb ut in exc^Sy protesUf aessum, ft Benjamin obsoleto. 
Hunc sonum coiTeptum Angli semper exprimunt per e brevem; ft e 
brevem nunquam aliter pronunciant nisi ante r, ubi propter tremulam 
ipsius motionem, ft vocalis subtilitatem subiti correptione comitatam, 
vix aliter efferri |)otest quam ur ; ideo per in pertain pertineo, ft pur 
in purpose propositum ejusdem sunt valoris. Vera hujusce soni pro- 
ductio scribitur per a, atque a longum falso denominatur ; ut in eane 
canna, wane deflecto ; & ante ge ut age setas ; in cseteris autem voca- 
bulis, (m faUor) omnibus ubi e quiescens ad finem syllabse post a, 
adjicitur ; u gutturalis . . inseritur post a ; ut in name nomen, quasi 
scriberetur na-um dissyllabum.' He proceeds to say that this sound 
is usually written ai or ay, sometimes «y, and rarely ea. 

'Post a in omnibus, nisi in eane canna, tvane deflecto, stranger 
advena, strange alienus, manger prsesepe, mangy scabiosus, ft ante 
ge; ut age eetas; inseritur u gutturalis, quae nihil aliud est quim 
continuatio nudi murmuris postquam a formatur, nam propter exili- 
tatem, ni accuratius attenditur ; ad proximam consonantem, sine 
interveniente u non facile transibit lingua. Differentia auribus, quss 
sonos distinguere poasunt, manifestb apparebit in exemplis sequenti 
ordlne dispositis. 



abreyis. 
B&r vectis 
hhb effutio 
cap pileum 
car carrus 
ca^ catus 
dsish allido 
Jleali fiilguro 
gtish csesura 
graand grandis 
land terra 
m^sah farrago 
jjsJt aptus 
for pix fluida 



aloDga. 
Barge navicula 
hlsat flatus 
carking anxietas 
carp carpo 
east jactus 
(iarfjaculum 
Jlasket corbis genus 
gasp oscito 
grant concedo 
lanch solvo 
tnaa^larra 
path semita 
t&ri Bcriblita 



Bare nudus 
blazon diyulgo 
ea^capa 
care cura 
case theca 
date dactyluB 
Jlake flocculus 
gatejasmak 
grange villa 
lane vioulus 
mason lapidarius 
/Kite caput 
tares IoMa 



'Si quid amplius ad hanc veritatem confirmandam velles, accipe 



214 HISTORY OF ENGLISH SOUNDS. 

ezempla sequentia ; in quibus at leniter pronunciata sonum habet 
a puree ; ut in cane, a verb post se admittit u gntturalem ut, 
Bain balneum Hail grando Maid virgo 

bane yenenum hale traho made fsLctus 

main magnus lay*n jacui ;;atn dolor 

mane juba lane yiculus pa/ne quadra 

plain manifeBtuB spaid castratus tail cauda 

plane laevigo spade ligo tale fabula.' 

778. Miege (1688) says : 

' Dans la langue Anglaise cette yoyelle A B*appelle et se prononce 
at. LorB qu'elle est jointe avec d'autres Lettres, elle retient ce m^me 
Son dans la plupart des Mots ; mais il se prononce tantdt long, 
tantdt bref. L'a se prononce en ai long generalement lorsqu'il est 
Buivi immediatement d'une consonne, et d'une e final, Exemple fare, 
tare, care, grace, /able^ qui se prononcent ainsi, faire, taire, caire, 

graice,faihle D'ailleurs, a se prononce en ai bref ou en e ouvert, 

lorsqu'il se trouve entre deux Consonnes, au milieu des MonosyllabeB ; 
comme hat, cap, mad. Mais il approcbe du Son de n6tre a, & la 
fin des NomB en al, ar, & ard qui ont plus d'une syllabe. Exemple 
general, special, cmimal, Grammar, altar, singular, particular; 
mustard, custard, bastard, vizard, & autres semblables. Excepts 
regard, qui se prononce regai/rd ; award & reward oil il Bonne comme 

en Fran9aiB Dans le mot de Jane Va se prononce on e 

masculin, Dghie* 

Fr ai had by this time evidently been smoothed into {e), so 
Mg's account fully harmonises with the other evidence. 

779. Mg, like W., makes no qualitative distinction between 
short and long a. Cp, on the other hand, expressly states 
that the vowel in such words as wane was not the long of x in 
cat etc, but of « in ken, which would make it either \k or [«. 
He finds this pure {ee) in the words cane and wane, in a before 
(nds) and (d^), as in strange, age, and in ai, as in tail : (k^^n, 
vreeri, str^md;, eeA^S, UeY), Elsewhere a vowel-murmur is 
added, as in name (n^^m), tale (t^^l) distinguished from (t^^l) 
^taU. The pure {ee) is evidently due to the influence of the 
front i vDl ai and the once front cons. (ds). This distinction 
between {ee) and {eed) cannot, however, have been kept up 
long, for there is no trace of it in LE, in which Cp's two 
vowels are both represented by (ei), as in (teil)=toi^, tale. 



MODERN ENGLISH : VOWELS. 2 1 5 

780. The vowel in all these words is ME a or ai (in some 
cases au), Eut Cp also recognises a lengthening of his (8e)= 
ME a, as in caf^? (keeserp) contrasted with car (}LBdT)y path (psese)?) 
with pat (psBt). His examples are not enough to enable us to 
determine with certainty the conditions of this lengthening, 
but it seems to have been regular before r and 9 followed by 
another cons, (kseserp, dsesert ; kseaest, gsesesp), the short vowel 
being preserved before single r and 9h (kser, dae/). He has 
(se^) before (}?) in (p8e8e)>), and the analogy of L£ would 
make us expect the same lengthening before single s in glasf 
etc. This distinction between ' Modem-long' (sese) and ' Mid- 
dle-long' {ee^ eed) is borne out by LE, in which (paa}^, peit) 
correspond to Cp's (paesa)', p^e^at), (aeae) having been broadened 
into (aa), while (sb) remains unchanged, as in (pset). This 
Mn-long (sese) is not found in the fMn authorities, who write 
(a) in (kast) etc. It must, indeed, have come in after the 
change of fMn (8e8e)=ME a into {e€\ for otherwise we should 
have had ^(pei]?)=jDa^i in LE. Hence, on the other hand, if, 
as was highly probable, path had (sese) in W.'s pronunciation 
as well as Cp's, his mane cannot possibly have had (aeae) also, 
but must have had Cp's {ee) or (eeo). It now appears probable 
that Bt's distinction between a and a (774) may, after all, 
have been identical with Cp's — that he, too, pronounced 
(msen, Tneen or me^n). 

781. The three sounds (ae, seee, ee)y as in man^ path, name^ 
were preserved unchanged in thMn, except that (ee) was per- 
haps narrowed to (ee) [♦ towards the end of the i8th cent. 
In LE they are represented resp. by (se, aa, ei) x» l^l) C^-C-^* 
The present (aa) is still ignored by Sheridan in 1780, who 
only admits (^Bsseij^) etc. 

782. The LE (aa) corresponds not only to Cp's (sbsb), but 
also to his (ee) when followed by r, so that his (kssr) and 
(kseserp) both have the same vowel (kaar, kaap), which is, 
however, of totally different origin in the two words. In 
(kjBOk'pj—cafp it is the result of an isolative change of (aeee) 
into (aa); in (kaar)=caf the change is combinative — due to 
qualitative influence of r which is so marked in thMn and 
LE (904). The stages of the latter change were (kasr, kar, kaer, 



2l6 HISTORY OP ENGLISH SOUNDS. 

kaax), the r bemg then dropped when not followed by a 
vowel beginning the next word. eMn (ee) followed by f + a 
Towel in the same word is, in accordance with the general 
principle (905), kept unchanged, as in (nBeTO\i)=narr(na, 

788. It will be observed that LE (aa) always points to IME 
a (or e before r), never to a, which in LE is represented by 
(ei) everywhere except before r, where it retains the earlier 
(<?<?)-sound, as in (he9r)=ME Adre, contrasted with (haad)=ME 
Aard. Exceptions are only apparent. Thus LE (raaiSar) 
points to a ME ropery the ME doublet roper being represented 
by the archaic vg (reiiSar). (aar), again, corresponds not to 
the strong ME are, but to the weak are, or, the original strong 
form being represented by the vg (ear). 

784. Li fMn an «-glide was developed between a and l, as 
shown by Td's occasional spellings fouU, caulfe='ilLE faUe^ 
calf. Salesbury says that in the English calme, call, the a * is 
thought to decline toward the sound of the diphthong au^ 
And again : * Sometimes a has the sounde of the diphthong 
aio especially when it precedes I or II, as may be more clearly 
seen in these words: balde hawld calvus, ball, howl, pila; 
WALL wawl murus.' Li the next cent, this diphthong generally 
followed the fate of ai«=ME an, being smoothed into j«, 
whence our j* in (fol) etc (859). For the an in eMn a{ti)rmcer 
etc see 860. 

785. In LE a and ar are rounded to (o) and (0) resp. after ir, 
as in (swolou, woz, wont, whot) =««7a//(?«7, wa;^, want, what, (dwof, 
Bwom, wop)=rftt7af/, swarm, warp» Also after tt?r=eMn 0)0 (919) 
in (roj?, T^}^)^ wrath and the vg (rop)=«TOj5 (reep). IME a is 
not affected in this way, wavien, wdnieti appearing as (weiv, 
wein) wove, wane in LE. (wotor) seems, together with (roj?), to 
point to sMn (waeseter), although the fMn authorities write 
the word with (aa). The first recognition of this influence 
of tt? is in Cp's statement that the a in was is a ' guttural ' 0, 
Wk still writes (wsez) in his phonetic transcription. Even 
in LE the rounding is barred by a back cons, following the 
vowel, as in (weeks, WBBg, tweeg) wax, wag, twang. Of course, 
where the combination (wo) is the result of parasiting before 
I, as in (wok)=«?ri(«)/^, a following back cons, has no effect. 



MODEBN ENGLISH: VOWELS. 21 7 



786. The Welsh HVg and the transcriptions of Sb express 
E. i in some words by i, in others by y, not at random, but 
according to strict rules, i is used to express final weak i, 

I before the back nasal in n^y ni, before cAt (^^^)=oo, and 

once before sA. y is written before the foreward conss thy *, 
fi, ty and the lip conss. Vy m. Of the following examples 
those that occur only in HVg are marked H, those which 
occur only in Sb are marked S, those which occur in both 
being left unmarked : 

-i: ladi, michti Hy redi H, 

-ing : king, thing Hy wynning II, blessing Hy gelding Sy 
Pegging & 

-ink : wrinkl 8y twinkl 8. 

-icht : richt, knicht, micht H, bricht H. 
g -ish : wish H. 

-yth: wyth H. 

-ye : ddys, ys iT, hys Hy blys(s) H, thystl 8. 

•yn : wynn, yn H, syn if, thynn 8, 

-yt : y t ff. 

-yf : lyf * live ' H. 

-ym : hym ff. 

Exceptions are very few. is for ys occurs once in HVg, and 
is no doubt a mere scribal error, as also Ady in Sb against Aoliy 
lili. The first i of the latter word may be due to a syllabifica- 
tion U4i, Unfortunately there are no examples before k and 
gy where the analogy of the nasals would lead us to expect «. 

787. In North Welsh % is narrow I both long and short, 
and ^ is T in some positions, j in others, Welshmen tending to 
identify our short f with their T. There can, therefore, be no 
doubt that in fMn i had the two sounds I and X, the original 
I being preserved before back and front conss., including the 
once fronted (/), and also finally. It is this narrow I which 
is probably indicated by the frequent eMn ie for y in ladie 
etc, and by the later ee in coffee etc, which could not well have 



2l8 raSTOEY OP ENGLISH SOUNDS. 

had long (ii). We now pronounce f everywhere, in lady as 
well as king etc. When the change took place before conss. it 
is impossible to determinate accurately. The widening must, 
however, have been completed in Cp's time, as shown by his 
pairing meet and need as short and long (777). 

e 

788. e is now [ in the South, \ in the North of E. and in 
Scotland. The eMn sound was probably [, for \ would have 
either remained unchanged, or been widened to X) or raised 
to [ — a sound which, as we see from Cp's vowel-scale, did not 
exist in sMn (777). 

780. The change of e into a before r, which was already 
carried out in IME in such words as Afln«i= older hergien^ 
Aarvest =^ihe normal iervest, is carried much further in MnE. 
er + vowel is preserved unchanged, except, of course, in Aarry 
and tarry, where the a is IME. All final ers, on the other 
hand, become ar in fMn, with the exception of the weak Aer, 
the older e only occurring in the earliest texts, the change 
being also very general when a cons, follows the r : star (starre 
Td, sterr Cheke), far {farre Td)=ME sterre, fer; marsA, starve, 
dark (a, e Td) = ME mersA, sterven, derk. darling, fartAing point 
to shortness of the ^ in ME derling, fertAing. Many of the 
words in which the change has not taken place are written 
with ea, pointing to group-lengthening : earl, earn etc. Several 
of these, however, have (aa) in LE, such as AeartA, Aeart (Aerte 
Td). As there seems no reason to suppose that the vowel of 
Aerte was ever lengthened, the spelling with ea may be a mere 
orthographic compromise between Aert and Aart, which last is 
the spelling of the phonetically identical equivalent of ME 
Aert * stag.' 

700. The change of (e) into (83) in the LE tAresA, tArasA is 
due to the preceding r, which has an analogous efiect in break 
(821) and broad (841). 

u 

701. Sb transcribes short u with his Welsh «r ), as in Iwst, 
block, gwt. The HVg also has w in most words : fwl * full,' ws. 
Atoning, swking. But, remarkably enough, it has the same 



MODBEN ENGLISH: VOWELS. 219 

symbol as it employs for f (786) in some words, nl jr. In synn 
= 0E 9unu^ sunne the y is constant. It occurs also in yntio 
* unto/ sym (miswritten ^nn), trysti, lyf * love,' syU * such,' and 
in %yprest. In all of these cases, except the last, the analogy 
of % y would lead us to expect a wide i. The y of 9yi9 might 
be explained bjb the representation of a dialectal u (800), but 
this will not apply to ^», and the conclusion seems inevit- 
able that the writer of HVg meant to indicate a distinction 
between 1 and t. As i has a lower pitch than 1, and con- 
sequently sounds less markedly rounded, and nearer the ob- 
scure (o), it was natural to denote it by the Welsh y in its 
second value j. That this symboUzation was not carried out 
strictly, is evidently due to the ambiguity of using y in the 
two very different values f and 1. HVg also writes y for 
shortened b = ME in dynn * done,' the spelling dywn being 
apparently a compromise between dyn and dwn. The latter 
would be the more correct spelling, as the vowel was pro- 
bably narrow. 

792. The other fMn authorities give no indications of a 
distinction of wide and narrow (u) ; nor, indeed, do any of 
them give any dear account of the mechanism of the voweL 
Smith, however, pairs /i^// and/00/, to and tooy hut and hoot etc, 
as short and long, and Bi^itler says that while 00 and u long 
differ much in sound, ' when they are short they are aJl one ; 
for good and gud^ hlood and hlud^ woolf and wulf have the same 
sound.' As 00 had the sound \k in fMn (833), and ZAftdl, to^ 
taolf preserve the sound 1 to the present day, we may con- 
fidently assume that hut^ luck^ mud, whose u Smith also pairs 
with 00, had either 1 or ). The fact that Florio in 16 11 iden- 
tifies the E. t» in tun, Hud, dug with the Italian close }, points 
clearly to 1, not ), in these words at least. 

798. In the phonetic writings of the sMn period we fiild 
the first indications of the change of u into a sound resembling 
our present ] in hut. Wallis says : 

' U vocalis qaando corripitur effertur bodo obBCuro. Ut in hut 
fled, cfui seco, 6ur lappa, hunt ruptus, ewrti maledictuB, etc. Sonam 
hunc Qalli proferont in ultima sjllaba Tocis mrviteu^, Dififert k 
Galloram « feminine, non aliter qoam quod ore minus aperto 



220 HISTOEY OP ENGLISH SOUNDS. 

efTeratur. DiBcrimen hoc animadyertent Angli dam pronunciant 
voces Latinas iter, itur ; ter ter, turiur-; cerdo siMrdo ; temtis Tur- 
nus; terria turris ; re/ertumy/wiimi, &c/ 

In his theoretical part he gives the following farther 
particulars of the French efosmininum and the u obicurum: 

^ Eodem loco/ that is, in sum/mo giUture, ' eed apertura faucium 
mediocri/ that is^ less than for d, ' formatur Gallorum e foemininum ; 
Bono nempe obscuro. Non aliter ipsius formatio differt k formatione 
prsecedentiB d aperti {p)i quam quod magis contrahuntur fieLuces, 
miniis autem quam in formatione Yocalis sequentis (i). Hunc 
Bonum Angli vix uspiam agnoscunt; nisi cum vocalis e brevis im- 
mediate praecedat literam r (atque hoc quidem non tarn quia debeat 
sic efFerri, sed quia vix commode possit aliter; licet enim, si citra 
molestiam fieri possit, etiam illic sono vivido, hoc est, masculo, 
efferre ;) ut vertite vii*tus, liberty libertas <&c. 

' Ibidem etiam, sed Minori adhuc faucium apertur4 sonatur d 
vel H obscurum. Differt k Gallorum e foeminino non aliter qu^m 
qubd ore minus aperto, labia proprius accedant. Eundem sonum 
fer^ efferunt Galli in postrema syllaba vocum aerviieur^ aacrificateur, 
etc. Angli plerumque exprimunt per H breue, in turn, verto ; hwm, 
uro ; dvU, segnis, obtusus ; cut, seco, etc. Nonnunquam o et ou 
negligentius pronuntiantes eodem Bono efferunt, ut in come, veuio; 
adme, aliquis ; ddne, actum ; cdmpani/, consortium ; country, rus ; 
couple, par; cdvet, concupisco; Idve, amo, aliisque aliquot; qu8d 
alio tamen sono rectius efferri deberent. Cambro-Britanni ubique 
per y scribunt ; nisi qubd banc literam in ultimis syllabis plerumque 
ut i efferant.' 

704. Wilkins describes short f» as ' a simple letter, apert, 
sonorous, guttural ; being framed by a free emission of the 
breath from the throat.' 

706. Holder has, as Mr. Ellis remarks, * very acutely an- 
ticipated Mr. M. Bell's separation of the labial and lingual 
passages, and the possibility of adding a labial passage to 
every lingual one.' He says : 

' In the larynx is depressed, or rather drawn back by contraction 
of the aspera arteria. And the tongue likewise is drawn back and 
curved ; and the throat more open to make a round passage : and ^ 

though the lips be not of necessity, yet the drawing them a little 



4 



MODERN ENGLISH: VOWELS. 221 

rounder, helps to accompliBb the pronunciation of it, which is not 
enough to denominate it a labial vowel, because it receives not 
its articulation from the lips. Oo seems to be made by a like 
posture of the tongue and Aroat with o but the larynx somewhat 
more depressed. And if at the same time the lips be contracted, 
and borne stiffly near together, then is made B ; u with the tongue 
in the posture of t but not so stifiP, and the lip borne near the upper 
lip by a strong tension of the muscles, and bearing upon it at either 
comer of the mouth.' 

'B is made by the throat and tongue and lip; in 8 the tongue 
being in the posture, which makes w> ; and in u in the same posture, 
which makes t, and in this 8 and u are peculiar, that they are 
framed by a double motion of organs, that of the lip, added to that of 
the tongue ; and yet either of them is a single letter, and not two, 
because the motions are at the same time, and not succeBsive, as are 
eu, /;/a Ac. Yet for this reason they seem not to be absolutely so 
simple vowels as the rest, because the voice passeth successively 
from the throat to the lips in 8 and from the palate to the lips in Uy 
being there first moulded into the figures of oo and t, before it be 
fully articulated by the lips. And yet either these two, 8 and t«, 
are to be admitted for single vowels, or else we must exclude the 
lips from being the organs of any single vowel since that the mouth 
being necessary to conduct the voice to the lips, will, according to 
the shape of its cavity, necessarily give the voice some particular 
affection of sound in its passage, before it come to the lipa; which 
will seem to make some such composition in any vowel which is 
labial. I have been inclined to think, that there is no labial vowel, 
but that the same affection from the lips may, somewhat in the nature 
of a consonant, be added to every of the vowels, but most subtlely 
and aptly to two of them, whose figures are in the extremes of 
aperture and situation, one being the closest and forwardest, which 
is t, and the other most open and backward ; there being reason to 
allow a vowel of like sound in the throat with B, but distinct from 
it as not being labial, which will be more familiar to our eye if it 
be written oo ; as in cu^ eooi^ fuU fad, tut toot, in which the lip 
does not concur; and this is that other. Thus u will be only % 
labial, and 8 will be oo labial, that is, by adding that motion of the 
under-lip, i will become u, and oo will become 8.' He proceeds to 
use his i, «, 8 in the formation of diphthongs and concludes thus : 
' Concerning 8 and u, this may be observed, that in subjoining 



222 HISTORY OF ENGLISH SOUNDS. 

them to another vowel, 8 is apter to follow a and o, hecause of 
their reeemblance in the posture of the tongue, as hath been said; 
and for the like reason u is apt«r to follow a and 6, as SaSl wawl ; 
efuge etc. But generally if the vowels follow, then it b 8 precedes 
and not u* 

796. Cooper says : 

' U formatur tantiim in gutture, k larynge spiritum vibrante, 
nudum efiiciente murmur, quod idem est cum gemitu hominis segritu- 
dine vel dolore ezcruciati ; quodque infantes (priusquam loqui 
valeant) primdm edunt: Et fundaraentum est, k quo omnes eoBteroe 
vocdUs, varii modificatione constituuntur .... Hunc sonum cor- 
reptum vix unquam aliter pronunciant Angli qu^m in nut nux; 
prout etiam in lingu4 latin4, ni ubi consonans prsecedens sit labialis, 
ut priiis dixi, et labiis dat formam qui sonus plenior efifertur, ut in 
pvU vello, inter hos minima datur, datur tamen specifica, diffe- 
rentia; ille etenim sonus dilutior est, hie plenior, ille formatur a 
larynge tantimi in gutture, hie k labiis contractis ; dum itaque o 
labiis formatur in sono continuato, si recedant labia in oblongam 
formam formatur u gutturalis ; in quibusdam scribitur per o ut, 
to come ^ venire ; Galli hoc modo, vel saltem persimili, olim sonarunt 
fsBminarum e, ut in providence, Germani syllabus Aam et herg in 
propriis nominibus. Nunquam in proprio sono apud nos productum 
audivi, ni in music4 modulatione, vel inter populos, prsecipu^ pueros 
cunctanter pronunciantes; pro longa enim vocali assumit dipthongum 
eu' 

707. Miege says that short u is pronounced o (meaning 
Fr o) in but, cut, rub, tip, under, run, eu \ji U9, 

708. As regards the formation of the sound, we learn 
nothing from Mg, and aU we learn from Wk is, that it was 
a back vowel. Wk's statement that it is formed * without 
any particular motion of the tongue and lips * would, if taken 
literally, point to some unrounded mixed vowel, such as \. 
This would be very near our ], but as W. and Holder agree in 
describing a very different sound, it seems safer to assume 
that Wk's statement is simply vague and inaccurate. W. 
states expressly that « is a back vowel of an obscure sound 
closely resembling the open Fr eu {♦, formed with a narrower 

^ come lA meant for the example, not tos (ta). 



MODEEN ENGU8H : VOWELS. 223 

jaw-opening than the E. e in vertue^ which he identifies with 
the Fr e feminine, this sound again being formed with a 
closer jaw-opening than the j* (now j*) oi fall. The Fr vowel 
in le is now a half-wide {, but as it is a weakening of e^ being 
actually identified with [ by the Lyons phonetician Meigret 
in the 16th cent., there seems every reason to suppose that it 
may once have been *[, as in Gm gahe^ which in some parts of 
South Germany is still [. Anyhow, it is certain that sMn e 
in her etc must have had some such sound as *[, for it is now 
X». W., then, distinguishes three back vowels according to 
height, his ^ back ' including ^ mixed ' as well : 



high-back 

mid-back 

low-back 



1 or 1 as in hut^ bur 

1 „ X „ „ Fr le, E. her 

J >i J j> >» /«^- 
Holder's description is equally clear, and can leave no doubt 

as to the value of his symbols : (70 = 1, 8 = 1, i^ = f . He does 
not state expressly, as W. does, that 1 was a high vowel, but 
he is quite explicit in identifying its tongue-position with 
that of 1. This much is therefore certain, that the first step 
in the change of ME u into the present ] was simple un- 
rounding without change of tongue-position. This unround- 
^g} Agfti^) was probably a gradual process, such as is now 
going on in Swedish, where original u is represented by I9, 
while u has become Ii>«. It is quite possible that this un- 
rounding had begun in fMn, which would make the identifi- 
cation of wide 1 = 1) with Welsh y still easier (791). There 
must, therefore, have been a time when the u of cut^ full was 
exactly half-way in sound between the (u)-group and the 
(9)-group. It is to this period that we must probably refer the 
present return of («) to the (u)-quality which has taken place 
regularly between a lip cons, and an /, which in E. has always 
had something of an (u)-timbre, as shown by the fMn change 
of (al) into (aul) etc. (u) = ME u is, accordingly, now fixed 
in (wul, wulf, ful, pul, bulek), appearing also in some words 
before other conss. when to precedes: (wumon), (wud) ^silva' 
against (wander) etc. But in the last two centuries there was 
considerable fluctuation, especially after «? as in (wumsen, 
wvmsen). Buchanan and Sheridan still have («) vx fvhame 



224 HISTOEY OF ENGLISH SOUNDS. 

against (u) in full^ and the tradition of (p'et, bi^Jer) etc has 
not yet died out, being partly maintained by the influence of 
Midland and Northern speakers. Holder, indeed, even seems 
to unround (uu) = ME o, as in fod^ \iirhich he makes the long 
of fuU^ in both of which he says * the lip does not concur.' 
But as Cp give (u) in fuU^ it seems probable that in Holder's 
pronunciation there was 9ome rounding ; as we see, W. ignores 
the undoubted rounding of j«, and Holder himself, while, 
admitting the rounding of (?, does not consider it essential. 
The utmost we can allow, therefore, is, that Holder's /a/!?, /(?<?/ 
were pronounced with 1), 1h respectively. But as oe? is still 
fully rounded in E., there is some difficulty in realizing such 
a pronunciation. According to Mr. Ellis, Holder makes a 
distinction between the vowels in fool and in two^ which all 
earlier and later authorities identify. It really seems safest 
to assume that Holder's u in but, full etc was so exactly half- 
way between 1 and 1 that ho was apt to confuse the sounds. 
This is confirmed by Op's statement that the difference be- 
tween the u in nut and pull though ' specific ' is * minute,' the 
two sounds differing only in the absence of rounding in the 
former, and its presence in the latter sound. When he says 
that becomes u in nut when unrounded, he probably means 
by the u of full, which he regards as the short of the }♦ of 
foal (777). Mg heard the E. « — as foreigners still do — as a 
sound between his Fr (? J ^^^ ^ { or ■£. 

700. We have no means of determining when 1 was lowered 
to its present representative ]. Certainly not as long as (u) 
and ("b) were confused, as they are by Holder, or continued to 
be almost identical, as Cp says they were. Hence we cannot 
accept Cp's identification of his u in nut with the natural or 
' ur-vocal,' which is repeated by Jones in 1701. Lediard in 
1725 still tends to confuse (u) and («), which latter he de- 
scribes as a quick, short German a j formed in the throat, 
which may be partly taken from Wallis, whom he follows 
closely. Of u in full he says that it has an obscure sound 
between German u short and E. ('b). Gm u short is now t in 
North Germany — Ld learnt his Gm in Hamburg — so this 
remark seems to show that Ld's n in full was not so fully 



. MODERN ENGLISH : VOWELS. 225 

rounded as in the present pronunciation, which is practically 
identical with the Gm. We may, therefore, assume that the 
complete separation of (u) and («) by the full re-rounding of 
the former and the lowering of the latter was not universally 
carried out till at least the middle of the i8th cent. 

u 

800. In conformity with its EMI origin, MnE generally 
has i for OE y^ as in itir^ sin^ hip. In some words ii became 
u in IME, as in mdcAe, KhE mucA (m«t|). Wherever we have 
(«) in LE, we may assume an intermediate (u), as in tcorry^ 
trust, 9ucA, A dialectal variety of U in ME was e, preserved 
in MnE merry, hemlock etc. Our (beri) = OE hehyrgan points to 
a different dialect from the written hury. In (bizi) and the 
vg (sitj) the (i) answers to the written u of hu9y and such, and 
in huild the spoken (i) = OE y is represented by ui, as also in 
guilt, though the u here is probably a mere sign of hardness of 
the ^, as in gueet. As there is no (u, ii)-pronunciation of huey 
and huild, their (i) is probably not a dialectal variation, but 
an unrounding of (y), which, as shown by the u, must have 
survived into MnE. We have direct evidence of such a 
survival. Sb says : 

'n vowel, answers to the power of the two Welsh letters u, w 
and its usual power is uw, as shewn in the following words tbub 
truw verus, vbbtux ffertuw probitas. And sometimes they give 
it its own proper sound and pronounce it like the Latins or like our 
own tv, as in the words bucke bu)ck dama mas, lust hoH libido. 
But it is seldom this vowel sound corresponds with the sound we 
give the same letter, but it does in some cases, as in bust htui, 
occupatns aut se immiscens.' Again in his pronunciation of Welsh 
he says : ' u written after this manner u, is a vowel and soundeth as 
the vulgar English trtLSt, bury, busy, ffuberden. But know well that 
it is neuer sounded in Welsh, as it is done in any of these two 
Englyshe wordes (notwythstanding the diaendtie of their sound) sure, 
ludce. Also the sound of u in French, or ii with two pricks over the 
heade in Duch, or the Scottish pronunciation of u alludeth somwhat 
nere vnto the sound of it in Welshe, thoughe yet none of them all, 
doeth so exactly (as I thynk) expresse it, as the Hebraick Kubuts 



226 HISTORY OF ENGLISH SOUNDS. 

doeth. For the Welsh u is none other thing, but a meane sounde 
betwyxte u and y beyng Latin vowels . . . and this vowell u alone 
amonge all the letters in Welsh, swarueth in sound from the true 
liatine pronunciation.' 

Welsh u is now T, a sound which, although quite distinct in 
formation from the Fr i, is so like it acoustically as to be 
identified with it by unpractised Welsh ears. As Sb himself 
states that his u is distinct from, though similar to, the Fr u 
and Gm «, there would be no difficulty in assuming that his 
Welsh u was the present T. But this is also one of the sounds 
of Welsh y — the vowels in «» dyn, for instance, being both 
I* or III. But the two vowels must have been distinct in Sb's 
time, for he identifies the Welsh y with the E. vowel in ^nne 
etc — that is, with X — and does not hint at any resemblance 
of Welsh y to Welsh or French u. He says that Welsh u lies 
between Latin u and y. As he says that Welsh u is the only 
Welsh vowel that diverges from the true Latin pronunciation, 
we may infer that to him the true Latin y was the Welsh T. 
As the other fMn authorities expressly contrast the * u Qal- 
licum ' with the * u Latinum ' = 1, which last Smith finds in 
the "E. fully book etc, there can be little doubt that Sb means to 
say that Welsh u lies between i and I or f . This would give 
the Norwegian I or Swedish I»4 as his «, which is, however, 
improbable. Another hypothesis is, that the two sounds were 
I and I, which last is near enough to X to justify Sb's identi- 
fication of Welsh and E. y. A third is, that W. y was X. 

801. Anyhow it is certaiil that Sb heard an f -like sound in 
the vulgar (vg London?) pronunciation of bury, bi/sy, trust, 
Huberden, and probably of other words as well. The first 
three words had u in ME, alternating with »', e and u according 
to dialect, and the last word contains the Fr Hubert, which of 
course had u. The ui of build ought to indicate a long vowel, 
and (byyld) is, in fact, one of the pronunciations given to 
this word by Gill. 



802. There can be no doubt that the eMn o was, like its 
ME predecessor, an open sound, for Florio identifies the Italian 



MODERN ENGLISH : VOWELS. 227 

open } with the E. vowel in god and dog, and Cp after him 
tacitly excludes short } from the list of E. sounds by confusing 
it with t, just as Florio did before him. Florio also identifies 
It. } with the long E. in stone etc, which was certainly }« at 
this time (839), Smith also pairing kop^ hope, Bullokar not, no 
as shorts and longs of the same sound. Gill, too, pairs to coll^ 
coale, writing them 0, <a resp., the latter being kept quite dis- 
tinct from the j« of tall, which he writes a and calls ' broad a.' 
It is, therefore, clear that fMn kept its ME sound }. 

803. In sMn W. and Cp agree in -pBxtvng folly fall as per- 
fect longs and shorts, showing that had now been lowered 
to the present j. W.'s account of the o-sounds is as follows : 

' d 6 aperta : Si apertura majori seu pleno rictu spiritus exeat, 
formatur Qermaiioniin d vel 6 apertum. Neque G^rmani Boliun sed 
et Galli, aliique non pauci, eodem sodo suum a plerumque proferunt. 
Angli sonum ilium correptum per 6 breue ; productum verb plerum- 
que per au vel aw, rarius per d ezprimunt. Nam in fdXl, foUy; 
hdU, hatd, holly; cdll, eoUar ; lawes, losse; cattae, coat; aw*d odd; 
9aw*d, 8od ; aliisque eimilibus ; idem prorsus Vocalium sonus auditur 
in primis syllabis, nisi qubd illic prodacatur his corripiator. 

' 6 rotundum. Majori labiorum apertura formatur 6 rotundum ; 
quo sono plerique proferunt Grsecorum ». Hoc sono GaUi plerum- 
que proferunt suum au. Angli ita fere semper proferunt pro- 
ductum yel etiam oa (ipso a nimirum nunc dierum quasi evanescente ; 
de quo idem hie judicium ferendum est ac suprjt de ea) : Ut, one, unus ; 
none, nullus ; whole, totus ; hole, foramen ; coal, carbo ; boai, cymba ; 
oat, avena ; thoee, illi ; chose, eHgi; etc. At ubi breve est, ut plurimum 
per d apertum (de quo supra) rarius per 6 rotundum pronunclatur. 

' Oo Bonatur ut Gkrmanorum t2 pingue, seu Gallorum ou. Ut in 
yocibuB good bonus, stood stabam, root radix, foot pes, loose laxus, 
loose laxo, amitto. 

' Konnunquam & ou negligentius pronunciantes eodem sono ' h H 
obscuro [«], * efferunt, ut in cdme, yenio ; sdme, aliquis ; ddne, actum ; 
edm2)any, consortium; country, tub; couple, par; covet, concupisco; Idve^ 
amo ; aliisque aliquot ; quae alio tamen sono rectius proferri debent.' 

804. Wilkins pairs folly, fall and full, fool, leaving foal etc 
without any corresponding short. Cp agrees in the first pair, 
but not in the second. He says of : 



228 HISTOET OP ENGLISH SOUNDS. 

* formatur k labiis paululdm contractis, dum spiriltia orhictUcUus 
emittur: ut in hope spes; productum semper, (nisi in paucis qu89 
per 00 [uu] sonantur; et ante I per ou [ou] labiales: at in hold 
audax) hoc modo pronunciant Angli, quern aliquando ecribont per 
oa; ut coach currus; correptus rar6 auditur, nisi in paucis, quse k 
consonante labiali incipiunt; ut post w in i(x>Z/^ lupus, wonder mirum; 
& in syllaba wor ; plura non memini : in quibusdam u hoc modo 
pronunciatur, ubi prsecedens yocalis est labialis ; ut pull, yello, fuU 
plenus; non quia debet, sed quoniam alitor facilius efferri nequit: 
Et 00 in good bonus, hood cucuUus, wood lignum; / stood steti; 
Galli per o ut globe globus, proteste protestor ; in copy exemplar 
corripitur. G&rmani per o, ut ostem pentecoste; quern in prin- 
cipio dictionum ferh producunt: in toort yerbum; Gott Deus cor- 
ripitur/ 

806. W.'s d rotundum and Cp's o cannot well be anything 
but }4, being rightly identified with the Fr au and the Gm 
long 0. The Gm short o is now J in North, } in South 
Germany: it is possible that the narrow sound prevailed 
formerly in North Germany also. W. says that short o is 
rarely pronounced with the short of }. It is uncertain whether 
he is here alluding to the older J in/o%, or is, like Cp, 
identifying the 1 of come with short }. 

806. Miege does not identify long o with the Fr au^ but 
with the Fr o, with which he also identifies the short o when 
not pronounced (9), adding ' il y a bien des mots ou T a un 
son mel^ de celui de Y a, et oil sans scrupule on le pent sonner 
comme un a,* which is, of course, a recognition of the very 
open sound of the £. short j. 

807. In L£ is lengthened to (oo) before the same conss. 
which lengthen (ae) into (aa) (780), as in (fro]?, kros, frost, of, 
ofii, soft) =/ro^A, crosSy frost, off, often, soft, the short (o) being 
kept before the voiced (z, v) in (gozlii), ov) = gosling, of, just as 
(se) is preserved in (sez). The short (0) is, however, still com- 
mon in (fro)?) etc, and some words never have (d) — only (0), 
such as moss, foster, gospel. The lengthening of (o) was no 
doubt contemporary with that of (se) ; Cp gives (oo) in frost 
with the remark * fere semper producitur ante st' W., on 
the other hand, does not yet acknowledge it, for he quotes 



MODERN ENGLISH: LONG VOWELS. 229 

loM and coit, now (los, kost), as containing short 0. For the 
development of (00) out of or see 905. 

808. In fMn a parasite-(u) was developed between and 
a following /, as in the case of a (784). Sb says : * before 
Id OT II i& pronounced as though to were inserted between 
them, thus colde, cawld frigidus, bolle batcl, tolle tawl 
vectigal.* 

Long Vowels : & 

800. The change of a through (seee, ee) into the present (ei, 
eo) has been described at length under a (781). It only 
remains to note that the main sources of Mn£ (sese) are new- 
long IME a, as in ndme^ and Fr a, as in blarney together with a 
few Northern as from OE and Scand. a, as in Aale^ race = OE 
hdly Olcel. rds. 



810. In HVg and Sb ME * is transcribed ei. Thus HVg 
has ei, abeiding, Krei^t^O'E ic (ME I) onbulung, CrUt, Sb has ei, 
ddein etc, at the same time reprobating the current E. pro- 
nunciation of Latin tibi as teibei. Smith, on the contrary, 
from his E. basis, considers (ei) to have been the real sound 
of Latin ? , saying 

' I Latina, quae per se prolata, apud nos tantum valet quantum 
Latine, ego, aut oevhis, aut ettam,* 

where the diphthongal pronunciation of his 7 is identified with 
that of aye. Hart says plainly that E. i is sounded ei. Gill 
blames him for expressing 7 with ei, he himself having a 
simple sign for long i, nl J, which he carefully distinguishes 
from ei and ei=z{Qei) ; he says : 

^ Differentia significationifl (quoad fieri potest, & sonus permittit) 
orthografia discemitur. Sic /. ego. ei oculus, H ita/ — *Nec c, 
BepiiiB prseponitnr i, dicimns enim hei, adhortantes aut laudantes, 
k ei ETE oculus, ei etiam, ita: vbi tamen sonus vocalis, exiguum 
distat ab illo qui auditur in 9jn tuus, & mjn meus.' — 'Communis 
dialectus aliquando est ambiguus. Audies enim 9a% aut 9ei they, 
illi.' — ' I, est tenuis, aut crassa : tenuis est breuis, aut longa : breuis 
sic notatur i, vt in sin sinne peccatum : longa sic i. vt in wn seeke 
yisus, a, um : crassa autem fere est diphthongus ei ; sed quia sono 



230 HI8T0EY OP ENGLISH SOUNDS. 

exilior paulb quam si difiPunderemur in e, retinebimus antiquum ilium 
et masculinum sonum .... eumque signabimus hoc charactere j. vt in 
^n siGNE signum. Omnium differentia est in toin winne vinco, win 
WESNB opinor, lojn wtne vinum.' 

He says of diphthongs generally : 

'Nee tamen in omnium diphthongorum elatione, utrique Tocali 
Bonus integer ubique constabit. Etenim Yocalis prsecedens ssepe- 
numero acutius sonare videtur, & clariiis; in at et ei, ita auroB 
implere, ut .t. subiungi sequius esset, quam ad latus adheerere/ 

alluding evidently to the Greek forms a, 77. We are here 
told that the j is nearly identical with the diphthong ei, in 
which, however, the first element is more *diflFuse'. This 
would imply that in j the first element is shorter or less 
distinct. G. himself says it is 'thinner* than in ei, which 
would point to some such distinction as [x, [l. But G.*s 
statements are so vague that all the certain result we get 
is that long i was a diphthong distinct from (ei) and (ai) 
or (sei), but closely resembling (ei). Now in Welsh the dis- 
tinction between ei and ai is very marked: *[!, 3^* ^^ ^» 
therefore, probable that Sb's identification of £. long i with 
Welsh ei really points to the present E. diphthong %£, and 
that this is also the value of G.*s j, although G. does not 
state distinctly that the first element was obscure in sound. 
The diphthonging of the old !♦ no doubt began with a partial 
lowering of the first half of the vowel, which would by further 
lowering develop either into [ or, with the help of a slight 
retraction, into \ : ItI, [■•■x, ^x. 

811. The descriptions of W. and Cp show that the fully 
developed sMn sound must have been the same as the present 
one. W. says: 

'I vocalis quoties brevis est sonatur plemmque (ut apud Gallos 
aliosque) exili Eono. Ut in bU morsus, loiU volo, s^l semper, win 
luoro, pin acicula, eUn peccatum, ftU impleo. At quoties longa est 
plemmque profertur ut Graecorum ci. Ut bite mordeo, wtU strata- 
gema, sttl0 stilus, fvtns vinum, ptns tabe consumer, etc., eodem fere 
modo quo Gallorum ai in vooibus main manus, j)ain panis, etc. 
nempe sonum habet compositum ex Gallorum d foeminino et i vel y.' 



MODBBN ENGLISH: LONG VOWELS. 23 1 

812. Cp says : 

' U in CvJt et t, diphthongum facillim^ constituunt, quam i longam 
Tocamus; at unne, vinum, hoc modo pronunciatur ante nd finales; 
ut blind cflBcus, vdnd ventuB : at pi7i'd pro pinned acicula subnexus ; 
k verbo to pin ; breyis est ; jnned marcidos ; k to pine marceo ; 
dipbthongus est. Scribitur per uiin. beguile fallo; disguise dissimulo; 
guide dax ; guidon Imperatoris baculus : per 01 in in-join in-juogo, 
joint junctura ; jointure dos, broil torreo, ointment unguentum.' 

This identification of the first element of the diphthong 
with the («) of cut need not, of course, be taken literally : it 
only means the first element was not (se) or (e), but some 
obscure vowel. 

813. The orthoepists of the i8th cent, agree generally with 
W. and Cp, although their analyses are often vague. The 
main divergence is that of Sheridan (followed by Enowles 
1847}, who sets up an Irish-E. pronunciation (oi), with the 
a of /ally the first element very short, and thus different from 
of, meaning perhaps only a broad jx, or, possibly, ]£, 

814. There was, however, another dialect of fMn which 
preserved the old i unchanged — or, at any rate, undiphthonged. 
Palsgrave and BuUokar ai^e the authorities for this pronuncia- 
tion. Pg says : 

' / in the frenche tong hath .ii. dyuerse manen of soundynges, the 
sonndyng of i, whiche is most generally vsed in the frenche tong, is like 
as the Italians sonnde t, and suche with vs as sounde the latin tong 
aright, whiche is almost as we sounde e in these words abee & flie, a 
beere for a deed corps, a peere, a felowe, a fee Sk rewarde, a little 
more soundynge towards t, as we sound % with vs.' 

' If t be the first letter in the frenche worde or the laste, he shall 
in those two places be sounded lyke as we do this letter y, in these 
words with vs, by and by, a spye, a Jlye, aiory, and suche other : in 
whiche places in those frenche bokes, as be diligently imprinted, 
they Yse to writte this letter y : but whether the frenche worde be 
written with i or y, in these two places he shall be sounded, as I have 
shewed here in this rule, as in ymage, conuerty, ydole^ estourdy, in 
whiche the y hath suche sounde, as we wolde give him in our tong.' 

< For as moche as v and i come often together in the French tonge, 



232 HISTORY OF ENGLISH SOUNDS. 

where as the v hath with them his distinct sounde, and the i is 
sounded shortly & confasely, which is the proprete of a diphthonge. 
I reken vi also among the diphthonges in the frenche tong, whiche 
whan they come together, shall haue suche a sounde in french 
wordes, as we gyue hym in these wordes in' our tong, a sioyne, I 
dwyne, I twyne, so that these wordes, agvysefy ctgvyU^ ccndvyre, 
dedvyre, (Mpourdhvy, meshvy, and all suche shall sounde theyr v and i 
shortly together, as we do in our tong in the words I have gyven 
example of, and nat echo of them distinctly hy himself, as we of our 
tong he inclined to sound them, whiche wolde rather say aviourdtwy^ 
ded^tf saufeondvyty gyuynge hoth to v and % theyr distinct sounde, 
than to sounde them as the frenche men do in dede, which say 
aviourdhvy, dedvyt, 8cmfc<mdvyty soundyng them both shortly to- 
gether, and so of all suche other.' 

816. The object of this last passa^ is to warn Englishmen 
against pronouncing Fr tii as dissyllabic fl instead of as a 
diphthongic fl, which monosyllabic pronunciation Pg ex- 
emplifies roughly by the E. stcyne^ although here, of course, 
it is not the i which is made into a glide, but the to. The 
important point is his distinct identification of long E. in 
Jjr, 9wyne etc with Fr I. His retention of the MTC sound 
is made a-priori probable, or at least possible, by the fact 
that in his pronunciation ME e in bee etc— all his examples 
are of ME ^, not f — had not yet become full f ♦, as was else- 
where the case in fMn (818), but had only got as far as 
a very dose (Danish) [-4, a sound between [♦ and I*. It must, 
however, be noted that Pg identifies the E, sound only with 
the initial and final Fr i, implying that the medial Fr i was 
not identical with the E. long i. This reservation, taken in 
connection with his statement about u (827) makes it possible 
that his long i was, after all, not absolutely identical with I«, 
being, perhaps, a slightly diphthonged sound — f-rX. K so, the 
pure !♦ was wanting in his sound-system. 

816. Bullokar says : 

'I, hath two soundes, the one agreeing to his olde & continued 
name, and is then a vowell, the other sounde agreeing to the old 
name of g, and of my ^ [d5], and is then a consonant.' 

He gives as examples : 



MODERN ENGLISH : LONG VOWELS. 233 

' I ly in my Bisterz kitchen witb a pillo'w bes/d her peticot, and 
thy wh/t pilion,* 

where the accent denotes length. He has no other distinc- 
tion between long and short i but the accent. He says of e : 

' e hath two sonndes, and vowels both, the one flat, agr^ing to his 
old and continued name: and the other sounde more sharpe and 
betwene the old sound of the old name of :e: and the name of :i: 
for such difference the best writers did use :ea: for :e: flat and long : 
& M, t>, 60 for :e: sharpe.' 

This statement is identical with Pg's, pointing clearly both to 
[^=ME e and I«=ME i, for there would be no sense in saying 
that [-4 (or !♦) lay between [♦ and Xi or any other diphthong ; 
we must, therefore, assume that Bll agreed with Pg in pre- 
serving ME i unchanged, or nearly so. 

817. The ME sounds e and f, e are in MnE distinguished as 
ee and e, ea : hed^ meet ; Aeal^ meaty mete—0^ hela {w\gemetan {£) ; 
halatiy ME mSte (OE mgte\ meten (OE metan). Final ee is 
shortened to « in subordinate words, as in he^ me, which are 
often written iee^ mee in eMn ; we still write ee in the less 
familiar thee^ partly to distinguish it from the. This MnE ea 
(as also the parallel oa) is probably a purely phonetic speUing, 
the a being added to indicate the opener sound. It occurs, 
however, at least once in the EUesmere ms of Ch {teare 
'lacrima'), and several times in TM: ckeape^ peaeee. It is, 
therefore, possible that it was suggested by some tradition of 
the eSth spellings heaved etc. The ME ie was also employed to 
denote the closer sound, as in believe^Jield^OAngl. gelefan^feld. 
In the earliest fMn books ea is hardly used at all. Caxton, 
who often writes ie for ME ^, does not employ it, and Mr. Ellis 
notes that even in Palsgrave's text (1530) it is very rare, 
though he employs it freely in his vocabularies. Tindal is, 
as usual, in advance of his time in his extensive use of ea^ 
although he is irregular, as the following examples will show: 

e : ye pr», se rd, fie, sene, slepe. deed, need, feale, deades. 

f : bred, est, este. yee ' yea,' see el ; breede. greate, meate. 
yer * ere* adv^ biestes. 



234 HISTOBY or ENGLISH SOUNDS. 

It will be observed that Td regularly assigns ee final to -f , 
in direct opposition to the later usage. His constant i€=f in 
bieH 'beast* is also a remarkable divergence from the later 
usage. His ea in J^a/;?='deed' may be a dialectal remi- 
niscence of WS dady though AR etc have did. Td has the 
usual MnE ea in heare^ dearer 01& geheran^ deorey in which the 
i was probably broadened by the r. 

818. Pg says : 

' £ in the frenche tong hath thre dyverse sowndes, for sonityme they 
Bownde hym lyke as we do in our tonge in these words, a beere, a beest, 
a peere, a beene and euche lyke .... The Bowndyng of e, whiche is most 
generally kepte with them, is snche as we gyve to 6 in our tong 
in these wordes ahoue rehersed, that is to say, lyke as the Italianes 
sounde e, or they with vs that pronounce the latine tonge aright: 
80 that e in frenche hath neuer such a sownde as we vse to gyue hym 
in these wordes, a bee suche as maketh honny, a beere to lay a deed 
corps on, a peers a make or felowe, and as we sounde dyuers of our 
pronownes eiidynge in e, as we, me, the, he, she, and suche lyke, for 
suche a kynde of soundynge both in frenche and latine, is allmoste the 
ryght pronunciation of i, as shall here after appere.' 

This passage, taken in connection with those already cited 
from Fg himself and BU (814, 816), is a clear statement that 
ME e in such words as he, the, she, we, me, bee, bier, peer had the 
very close, t-like sound [-4, while ME § in such words as 
bear 'ursus*, pear, beast, bean had an opener sound, which Pg 
compares to the Fr and Ital. e. He does not tell us whether 
these words had the sound of the close Ital. and Fr [, or 
of the open (e^)=p or ([♦. In the absence of any direct 
evidence, we may assume that ME f kept its open sound 
in fMn. 

810. HVg and Sb express the two sounds by Welsh i I 
and e J respectively. Thus HVg has toi, wiri, kmn, dids ; 
leving, leding, Sb has tsis 'cheese', kwin\ efer, bred * bread*. 
As Welsh has no [, the e does not point necessarily to (ee), 
or the i to I« in E. As regards we etc it is, indeed, possible 
that the Welsh i may mean [*♦, as in Pg's and BU's pro- 
nunciation. 

820. All the other authorities agree in pairing win, ween 



MODEEN ENGLISH: LONG VOWELS. 235 

etc as containing the long and short of the same vowel. 
As soon as the long i of wine had become a distinct diphthong, 
the dose (ee) of ween was moved up into its place, giving 
(wiin), a pronunciation which has lasted almost up to the 
present day, and of which our sl-rX^l is but a slight modi- 
fication. 

821. The narrowing of (ee) into [♦ would naturally follow 
the disappearance of [♦*. W. says : 

' 6 profertur sono acute claroque ut Oallorum ^ masculinum,' except 
before r ; ^ea effertur nunc dierum ut S longum : bodo ipsius a 
penitus suppresso, et sono literse e producto. Nempe illud solum 
praestat a ut sjllaba reputetur longa. Ita met obviam factus, meat 
victus, set sisto, sedere facio, seat sella, etc., non sono differunt nisi 
quod Yocalis illio correpta, hie producta intelligatur.' 

Here the statement that met, meat etc differ only in 
length must not be taken too literally, for W.'s main object 
evidently is to impress on his readers that the a in the latter 
word is simply a mark of length. The expressions ^ sharp ' 
and ^ clear*, and the comparison with Fr /, which is repeated 
by Mg, point distinctly to narrow (ee), which W. strictly 
separates from the a of maney this latter sound having itself 
become an open (ee) before W.*s time (780). All doubt is 
removed by Cp's pairing of mil and weal ©[tCD, ien and cane 
ax*! or a|)n. It appears from Cp's lists that in sMn ME i 
was regularly represented by close (ee), as in treal^ wean, breaks 
remaining open (ee) before r (with some exceptions) as in bear, 
early earn, (ee) also, according to Op, in scream, where it is due 
to the preceding r. The other authorities do not make this 
distinction of (ee) and (ee), so their (ee) is ambiguous as 
regards narrowness. 

822. Towards the middle of the i8th cent, the sMn (ee) 
became (ii), not only in sea, heal etc^ but also, in the mouths 
of many speakers, in such words as break, great, which are 
now always pronounced with (ei) = sMn (ee), which was 
preserved by the preceding r. 

823. There is a certain fluctuation between (ee) and (ii) in 
eMn. The ea before r in hear, weary, fear, dear no doubt at 



2^6 HISTORY OF ENGLISH SOUNDS, 



first indicated a real broadening, but this cannot have been 
general, for these words have (ii) assigned to them by nu- 
merous eMn authorities, the spelling ea being probably kept 
up partly to distinguish such pairs as dear, deer=ME dere^ 
defy fear^ /<?r<?=ME fer^ fere^ Jiear^ here=W£k 'here{n\ her. The 
spellings /^^, here instead of *feer^ *heer seem to point to occa- 
sional broadening in these words also. One result of this 
confusion between fMn (eer) and (iir) was that many -ear 
words with ea=ME § took the sound of (iir) in sMn, such as 
smear ^ near, tear sb. Otherwise (ii)=ME ^ is rare in eMn, 
the chief instances being evil and even. 

824. In many words, especially before the stop rf, ME § was 
shortened to « in fMn, and, of course, remained unchanged 
in the later periods : healthy breathy heavy, head, bread, breadth^ 
threat. So also ME e, especially before t, as in let, wet, 

825. There is a curious passage in Gill, from which it 
appears that the 1 8th cent, (ii) for (ee) had already developed 
itself in fMn, but only as an effeminate affectation. After 
observing that the eastern English are fond of thinning their 
words, saying (fir, kiver, deans) for (feier, kuver, dans), fire, 
cover, dance, he goes on to say : 

* la-xf^Tfira autem illam magnopere aflFectant trvyooroXoi nostrsB 
Mopsae quae quidem ita omnia attenuant, vt a et o, non aliter 
perhorrescere videantur quam Appius Claudius z. sic enim nostrsB 
non emunt (loon) lavm, et (kaambrik) cambric, Bindonis species ; sed 
(leen) et (keembrik) ; nee edunt (kaapn) capon caponem, sed (keepn) 
et ferfe (kiipn) ; nee unqnam (bntjerz meet) butchers meat camem 
k laniJB, sed (bitjerz miit). Et qunm sunt omnes (d5intlimin) non 
(d5entlwimen) gentlewomen, i.e. matronee nobiles, nee maids anciUas 
vocant (maidz) sed (meedz). Quod autem dixi de a, recanto ; nam 
si quando 6 gravistrepum audiretur, locum concedunt ipsi a, sic enim 
aliquoties ad me pippiunt (oi pre ja gii jar skalerz liiv ta plee) pro 
(oi prai jou giv juur skolars leev tu plai), / pray you give yov/r 
scholars leave to play, Qu8bso concede tuis discipulis veniam 
ludendi,* 

Such a pronunciation as (miit) for (meet) =»itffl^ would pro- 
bably, as Mr. Ellis observes, have sounded as affected to 
Cooper and his contemporaries as it did to Oill himself. 



MODERN ENGLISH : LONG VOWELS. 237 



826. ow=ME u is transcribed by the Welsh ow^ in HVg 
and Sb. Thus the HYg has note, ototy down^ owt=OK nu, ure^ 
ofduncy uty and Sb has now, ddaw, Cheke, Smith, Hart, and 
Oil! also analyse the sound as (ou). They all agree in making 
the first element short — (nou). The diphthonging of u is, 
therefore, quite parallel to that of (ii): h passed through 
i-ri into }i, which was afterwards diverged into (eu). Wallis 
says of the ou, ow in sow^ Aouse, out etc that it is pronounced 
with an obscure sound composed of obscure d or u (v) and lo, 
and Cp's description agrees (see the passages quoted in full 
885, 886). Lediard identifies the E. diphthong with the Gm 
au ]>. Sheridan analyses it into (ou), parallel to his (oi)=long 
f , meaning probably the same sound as Lediard. The present 
sound is x^^ 'with the first element lower than in Xi-r. The 
older pronunciation of sMn was probably 1^, of which jj£ and 
js- (still preserved in America) are independent developments. 

827. We would expect that dialect of fMn which preserved 
ME « as a monophthong — that of Pg and BU — also to have 
preserved u, Pg says : 

* Ov in the frenche tong shalbe sounded lyke as the Italians sounds 
this vowel v, or they with vs that sounde the latine tong aright, 
that is to say, almost as we sounde hyni in these wordes, a cowe, 
a mowe, a sowe, as 6vUre, sovdaj/n, oMier, and so ofsuche other.' 

We gather with certainty from this passage that ME 5 had 
not yet passed into its usual fMn sound (uu), and that the 
nearest approach to Fr h in Pg's pronunciation of K was 
the old u in coice=OE cu etc. If the 'almost' is to be taken 
literally, we can only infer — as in the case of * (815) — that 
old M had been very slightly diphthonged in Pg s pronun- 
ciation, ^i-ri. 

828. Bll says : 

' hath three soundes, and all of them vowels ; the one sound 
agr^ing to his olde and continued name, another sound, betw^ne 
the accustomed name of, 0, and the old name of, v, and the same 
sound long, for which they write 00, (as I do also, but giuing it a 
proper name, according to the sound thereof), the thirde sounde 



238 HISTORY OP ENGLISH SOUNDS. 

ie as, V, flat and short, that is to say, as this Billable ou, short 
sounded : for which some of the better learned did many times use, 
oOf Sc V, according to their sounds, but most times with superfluous 
letters/ 

He illustrates the three sounds by the words : 

(i) tonne filius, vpon, hosome (first vowel), come^ elote, 

(2) 9onne sol, out, bosome (second vowel), come. 

(3) ^^"^^, toke, boie, sone. 

^ U also hath thr^ soundes : The one of them a m^re consonant, 
the other two soundes, are both vowels: the one of these vowels 
hath a sbarpe sound, agreeing to his olde and continued name : the 
other is of flat sound, agreeing to the olde and continued sound of 
the diphthong : ou : but alwaies of short sounde.' 

Here, instead of pairing (u) in sun with the vowel of soon 
=:ME sdne, as Sm and the others do, he. puts the latter in a 
class by itself, and pairs (u) with the ou=ME u of out^ imply* 
ing that the latter was still 1« in his pronunciation. 

829. In room^ stoop, droop ^OE^ rUm, stupian, Olcel. drupa^ 
ME H has been preserved up to the present time (except that 
in the fii'st word the vowel is now generally shortened), evi- 
dently by the influence of the following lip-conss. The pre- 
servation of group-lengthened ME ii in the subst. wound may 
be due to the preceding w. The preservation of u before (}>) 
in youtA and uncouth is anomalous. 

U 

See under eu (861). 

830. The only native «-word preserved in MnE appears to 
be bruise=lVfS br^san (Angl. bresan). 



831. In MnE the ME and p, <^ are distinguished as 00 and 
0, oa: soon, stone, boatz=ME sone, st^, bpt, final 00, as in too, 
being sometimes shortened to 0, as in to, do, and oe being 
often written for final o(a), as in doe, toe against so, no. The 
digraph oa was evidently formed on the analogy of ea, for 
it came into general use later than the ea. It is rare in Td, 



MODERN ENGLISH: LONG VOWELS. 239 

who writes and 00 nearly at random, as in ME. The follow- 
ing are examples of his spellings : 

0: boke, sone. too jof^?, floore, good, bourd, bloud. shues. 

9 : holi, loth, soo, goo, go(o)Bt. moare, broade. 

832. The passages already cited from Pg and BU (827, 828) 
show that in their pronunciation ME had not yet been 
changed completely into (uu), as in the pronunciation of the 
other fMn authorities : Pg and BU probably pronounced hook 
exactly as the Swedish hok — d]-)*Q. 

833. The HVg and Sb identify the sound with Welsh w \. 
Thus HVg writes mwddyr^ gwd^W^i moder, ffod, and Sb writes 
tto *to, two', scwl, fficd=::ME fo, two, scdle, god. The other fMn 
authorities (except Pg and BU) agree, pairing /i<^ ajidfool etc 
as short and long (792). 

884. In sMn we find W. identifying E. 00 with the Gm 
long u and the Fr ou (803). That it was narrow, =):«, is made 
certain by Cp*s refusing to pair full and fool as long and 
short. It would be superfluous to prove that the sound h 
lasted through thMn, tiU it was diphthonged in the present 
cent, into l-ri*, lo, probably through i-ri*. 

835. fMn (uu) was shortened to (u) in some words, espe- 
ciaUy before (IS), (d) and cons.-groups. The shortening is, 
of course, oldest in those words which are written with u^ 
LE (n), such as gum, rudder ^ME gome, ro/^er, but we must 
assume an at least occasional fMn shortening in aU words 
with LE (11)= ME d, as in dolA, other, mother, done, flood, blood, 
month, monday. Td \i2A fludds:= floods. 

836. There is a further sMn shortening of (uu) to (u), which 
(u) is of course preserved in LE, the change of the earlier (u)s 
to (n) having been carried out before this new shortening 
began. With a few exceptions this shortening is general 
before stops, and occurs before other conss. also : (buzam, buk, 
fut, sivA)— bosom, book, foot, stood. The shortening in such 
words as (huf^ spun, rum)=^<^, spoon, room is stiU later, the 
long (uw) being stiU retained by many speakers. The shorten- 
ing before stops also was not general even in thMn, in which 
book still had (uu). But, on the other hand, we find (gud) 
as well as (guud) in fMn, although W. has only (uu). Hence 



24Q HISTORY OF ENGLISH SOUNDS. 

the sMn doublets (g^d, gad) etc, the first coming from fMn 
(gud), the second from fMn (gaud). These shortened sMn (u)s 
must have been narrow i at first. It is uncertain whether 
Cf'B/oot really means >io, or only >i*o with half-long vowel. 

9 

887. Palsgrave says : 

' in the frenche tong hath two diuers maners of Boundynges, the 
soundyng of o, wbiche is most generaU with them, is lyke as we 
sounde o in these words in our tonge a hoore, a soore, a coore, and 
Buche lyke, that is to say, like as the Italians sounde o, or they with 
YS that sounde the latin tong aright.' 

The last two examples show that the first word is meant for 
boarz^OE bar, all the words having ME p or d, 

888. Salesbury says : 

* in Welsh is sounded according to the right sounding of it in 
Latin : eyther else as the sound of o is in these Englyshe wordes : 
a Doe, a Boe, a Toe: and o never soundeth in Welsh as it doth in 
these wordes of Englysh : to, do, two,' 

And- again, speaking of English, he says : 

* takes the sound of [Welsh] o in some words, and in others the 
sound of w ; thus to, to, digitus pedis ; so, 80, sic ; two, tw, duo ; 
TO, <u? ad ; schole, acfol, schola .... But two oo together are sounded 
like iv in Welsh, as good gtvd bonus; poobe jtwr pauper.' 

Here the open E. o in toe^OE ^ etc is identified with the 
Welsh J4. The HVg has slso pop = pope (OE papa). 

839. Smith pairs as containing short and long 'o latina' 
the foUowing words, which are here given in their present 
spelling : 

Short : smock, horse, hop, sop, not, rob, hot, pop. 
Long : smoke, hoarse, hope, soap, note, robe, boat, pope. 

All the longs are ME g, as in wop, or ^, as in Aope. 

The others give similar pairs (802). Florio identifies the 
vowel in E. stone, tone, bone with the Italian open }. There 
can, therefore, be no doubt that ME p and S had the open 
sound (00) in fMn, which in the next period becomes }♦, 
pointing to fMn ^ rather than to j« or ]:«, one of which 



MODERN ENGLISH : DIPHTHONGS. 24 1 

(probably ihe former) was, besides, ihe usual fMn sound of 
ME au (856), which is still kept quite distinct from the of 
stone. It is not improbable that some fMn speakers made a 
distinction between ME p and ^, but we have no means of 
proving such a distinction. 

840. There is full evidence of the narrowness of the sMn 
in sUne etc (805), and this pronunciation continued down to 
the diphthonging in the present J^). 

841. In broad and in ihMnyroat we have (oo) corresponding 
to ME g, by the influence of the preceding r, parallel to the 
retention of (ei)=sMn (ee) in ^eat, 

842. The development of a para8ite-(u) between fMn (00) = 
ME g and a following I (cp 808) is shown in the spellings awld^ 
howld eic= oldy AM in HYg, and is confirmed by the other 
authorities. 

Diphthongs : ai, el 

848. The IME tendency to confuse ai and ei is observable 
in MnE orthography also, where ai, ay is written not only 
for ME ai, as in day, fair, nail, slain, maid, but also very 
generally for ME ^, as in way, sail, raise, rain, laid, for H in 
hay, betoray, and for §i in clay, stairs, ey, tf}=ME ei is still 
preserved in they, their. As the representative of ME H, pi 
it is more frequent ; grey ; key, whey, either, ei is always written 
before ghi neigh, neighbour, eight, weight. The spelling still 
varies in gray, grey. Td varies between ay and ey in graye, 
greye, rayne, reyne. He writes kaye throughout. 

844. The HYg has ei only in ddei, ddey. In all the other 
words it has only ai, ay, ae, as in aish ^ask', day, dae, away, 
awae, kae, agaynst, maedyn. Sb has no example of ei, writing 
vayne both for vein and vain in his E. examples, and tran- 
scribing it phonetically as vain. In the other words he tran- 
scribes with ay, as in nayl. This fluctuation between i, y, e 
aa the second element of the diphthong shows that it was 
not full X as in the Welsh ai ]r. Welsh ae is now ^, but the 
ay points rather to ]£ as the E. diphthong. 

845. Palsgrave's distinction between ai and ei is very 
clear : 

B 



242 HISTORY OF ENGLISH SOUNDS. 

'£% ymyersallj throagh out all the frenche tong ahalbe Bounded 
like as he is with vs in these wordes, obey, a aUy^ a grey, that is to 
say, the e shall have his distinct sounde, and the t to he sounded 
shortly and confusely, as conseil, tiermeil, and so of all suche other/ 

' At in the frenche tong is sounded lyke as we sounde ay in these 
wordes in our tong rayne, payne, fayne, diedayne, that is to say, a, 
distinctly and the t shortly & confusely/ 

846. Smith, on the other hand, Bays that the difference 
between them is minute, and that some words haye (ei) in 
the mouth of one speaker, (ai) in that of another, effeminate 
speakers substituting (ei) for (ai) generally (probably through 
IX): 

'Inter Ai Sc Ex diphthongos minima differentia est, prsesertim apud 
nostrates, apud nos tamen audiuntur hi soni. (Fein) fingere, (deinti) 
delicatus, (point) pingere, (feint) languidus. Bed non h»c tantum 
verha per ei pronuntiantur, sed caetera omnia per ai scripta mulier- 
culae queedam delicat lores, et nonnulli qui volunt isto modo videri 
loqui vrbanius per ei sonant, yt hsBC ipsa quee nos per ei scrihimus, 
alij sonant et pronuntiant per (u, tarn ahiaf^opoi sumus in his duntaxat 
duahus diphthongis Angli.' 

< Est diphthongus omnis sonus h duahus vocalihus conflatus ut : 
AI, (pai) solvere, (dai) dies, (wai) via, (mai) possum, (lai) ponere, 
(sai) dicere, (esai) tentare, (tail) cauda, (fail) deficere, (faain) libens 
ac volens, (pain) poena, (disdain) dedignor, (claim) vendico, (plai) 
ludere, (arai) vestire seu omare. In his est utraque litera hrevis 
apud vrhanius pronunciantes. Rustici utranque aut extremam 
saltem literam longam sonantes, pinguem quendam odiosum, et nimis 
adipatum sonum reddunt. (Paai) solvere, (daai) dies, (waai) via, 
(maai) possum, (laai) ponere. Sicut qui valde delicate voces has 
pronuntiant, mulierculee pnesertim, explicant plan^ Romanam diph- 
thongum ae, AE diphthongus Latina. Pae solvere, dae dies, v)ae 
via, mae possum, lae ponere.' ' Scoti et Transtrentani quidam Angli 
voces has per impropriam diphthongum Qrsecam a profenint ut nee 
i nee 6 nisi ohscurissime audiatur. A diphthongus impropri^ Grseca 
(paa, daa, waa, maa, laa).' 

By the * Latin diphthong ae ' Sm probably means {ee\ as it 
is not possible that he would note such a minute distinction 
8^ JXi ]c, and we know that ae was regarded as an ^-sound in 
the Middle Ages, being, indeed, often written e. 



MODERN ENGLISH : DIPHTHONGS, 243 

847. Gill (810) distinguishes (ei, eei) from (ai, aai), as in (%ei, 
'Seeir ; daai, wai, waai). 

848. Butler says : 

* The right sound of ai, au, ei, eu, ot, ou ; is the mixed sound of 
the two vowels, whereof they are made: as (halt, vaut, hei, heu, 
koi, kou) : no otherwise than it is in the Qreek/ 

' But ai in imitation of the French, is sometime corruptly sounded 
like 6 : as in may^ nat/j play, pray^ say, stay, fray, slay : specially 
in words originally French, as in pay, baili, travaU: though plaid 
have lost his natural orthography, and we write as we speak plead,* 

Here the coexistence of the two pronunciations (ai) or (sei) 
and the smoothing (ee) is clearly stated. 

In some pronunciations this smoothing had taken place 
much earlier. Hart in 1569 omits (ai) altogether from his 
list of diphthongs, and transcribes ai by (ee), for which he 
is severely blamed by Gill, writing fifty years later, who 
contrasts Hart's (ue, 'Se) with his own (wai, Sei). 

849. The diphthong sur\dyed even into the sMn period. 
Wallis says that ei, ey, were (ei) or even simply (ee) without 
the (i), but adds, * Nonnulli tamen plenius efferunt, acsi per ai 
Bcripta essent.' The diphthong ai he upholds still as a diph- 
thong, ^Ai vel ay sonum exprimunt compositum ex d Anglico 
(hoc est, exili) correpto, et y. Ut in voce day dies, praise laus.' 
This would give (daei) etc, which is also Wilkins's notation. 

850. Cp says : 

' Vera hujusce soni [yowel in ken] productio scribitur per a, atque 
a longum falsb denominatur, ut in eane canna .... hie sonus, quando 
pur^ sonatur,' that is when it is not {ee9), ' scribitur per ai vel ay ; 
ut pain dolor, day dies ; quae hoc mode in omnibus fere dictionibus 
plerumque pronanciantur : per ey in convey deporto, obey obedio, 
purvey rebus necessariis provideo, survey lustro, they illi, trey trulla, 
whey serum lactis: quandoque rarb autem per ea; ut pearl 
margarita. 

Corripiiiir in Producitnr in 

sell yendo saU nayigo 

sent missus saint sanctus 

R 2 



244 HISTORY OF ENGLISH SOUNDS. 

tell nuncio taU cauda 

teni tentorium tairU inficio.' 

This makes ai (ee) except in a few words. But afterwards 
he says: 

* At lenids prolata sonatnr ut a in cane ; fortius, plenum assumit 
Bonum diphthongi ai ; ut brain cerebrum, /rail iragOis ; ay finalis 
at a, sic day dies j at ante r scribitur pro a in affairs res, airy 
aereus, dairy lactarium, debonair candidus, despair despero, fair 
pulcher, fairy lamia, Judr crinis, pair par, repair reparo, stairs 

scala ; ceetera cum a/re ; ut are sunt, dare audeo Ai in 

bargain pactum, captain dux, certain certus, chapHain capellanus, 
curtain velum, forrain extraneus, fountain fons, moumtai/n mons, 
villain furcifer, & prior ai in mavrUain sonatur ut a correptum 
sive e breve.' Again he says: 'Sonus a in I can possum; I ea^t 
jacio; conjunctus cum t sonum literse ee exprimente; constituit 
dipthongum in bait esca; caitiff homo improbus; ay pro / vel yea 
imo ; k eight quam vulgariter pronunciamus ait, Plures baud scio/ 
^E in hen, vel a in Carts i propositus dipbthongum priori [sei] 
Bubtiliorem constituit ; ut praise laus : in paucis scribimus ei vel 
ey finalem; ut height altitudo; toeight pondus, & eorwey deporto, 
aliaque qu» supra sub e ostendimus; quibus exceptis ccetera 
scribuntur cum ai vel ay ut Juvinous detestabilis, plerunque autem 
in colloquio familiari, negligenter loquentes pronunciant ai prout a 
simpUcem in Cane,* 

The statement that ai is monophthongic when uttered gently, 
dissyllabic when uttered more strongly, seems to point to the 
existence of stress-doublets. There may have been a weak 
(Vee) corresponding to a strong (ieei) or ("Sei). 

851. In thMn ei and ai both settled down to (ee), which 
was perhaps narrowed to (ee) at the end of the period. 

852. We may sum up by saying that ai probably passed 
through (sei), (ei) into (ee), being in its third stage levelled 
under ei, ei must have had its first element open — (ri) — or 
else it would have been smoothed into (ee), not (ee). 

oi 

858. The E. oi, qy is transcribed oe in HYg in asoel^ and oy 
in Sb in tsioynt =: Joint, In Welsh oi = Jx, oe = Ji. Smith is 



MODEBK ENGLISH : DIPHTHONGS. 245 

doubifnl whether it should not be written ui, and Qill hesi- 
tates between (oi) and (uui) — ^where the doubling, as in (eei, 
aai), perhaps only indicates length of the glide — as in (boil, 
buuil). This change of (oi) into (ui) seems to show that the 
first element of the former was close } rather than }• 

854. Wallis, in the next century, says : 

' In ot . • . vel oy . , , praeponitor aliquando d apertum (at in 
Anglorum h6y puer, <6y« nugGB ....), aliqaando d obscurum, (at 
in Anglorum hdil coqueo, tdU labor, dil oleum . . .), quanquam non 
negem etiam horum nonnulla k quibusdam per apertum pro- 
nunciari.' 

Here we see the older (oi) retained, while the (u) of (ui) under- 
goes its regular change into (v). The resulting (vi) was then 
levelled under (oi), so that boil and bile, toil and tile etc were 
confounded, the oi being retained in writing. In the latter 
half of the i8th cent, the spelling caused a reaction against 
the pronunciations (boil, poizen) etc, which now survive only 
as vulgarisms, and the oi was restored. The analogy of the 
vb bail led also to the change of the sbst (boil) = OE byle into 
(boil), this being the only E. word of direct Omc origin which 
has (oi). 

an 

855. The E. au is transcribed aw in HVg, as in grawni, 
fawl 'fall' = Td's fanll (784), and in Sb, as in waw^waioe 
* wave,' watol * wall.* Welsh aw is J^. Sb says : 

'to English & to Welsh do not differ in sound, as wawe, toato 

unda, Also to is mute at the end of words in English, as in 

the following awb pronounced thus a terror.' He also 

says that * sometimes a has the sound of the diphthong ato especially 
when it precedes 2 or {2, as may be more clearly seen in these words 
BALDE, havold calvus, BALL, hawl pila, wall, wauH murus.' 

The pronunciation (aa) = awe is parallel to Sb's bo = bow (883), 
the dropping of the (u) being due to the length of the pre- 
ceding vowel. Sm, however, gives this word as (au). Sb 
himself in another place writes wifth aw, in which aw seems to 
be a phonetic representation of awe. Hart identifies E. au 



1 



246 HISTORY OF ENGLISH SOUNDS. 

with ibe Gm an. Bll, too, has diphthongic (au), against his 
smoothed at (848). 

856. Gill, on the other hand, who repudiates Hart's (ee) 
= aiy himself makes au a monophthong in most cases ; he 
says: 

'A, est tenuis, aut lata; tennis, ant brevis est, vt in (taloou) 
TAiiLOWB sebum; aut deducta, ut in (taal) talb fietbula aut com- 
putus : lata, vt in ^ talle procerus. Hunc sonum Germani 
exprimunt per aa, vt in maal conuiuium, Tuiar coma: nos ynico 
cbaractere, circumflexo d, contenti erimus.' 

* A prseponitur e, ut in cierj aebie aereus. o nunquam ; sadpius t, 
et u, yt, in aid auxilium ; bait esca ; latm sindonis species ; Sc a paun 
pignus : vbi aduerte cm nihil differre ab d. Eodem enim sono pro- 
ferimus a hdl, ball pila; et tu hdl, baule, vociferari: at ubi ver^ 
dipbthonguB est, a, deducitur in d, vt du awe imperium; duger 
terebra.' 

Here au is described as having the broad sound of Gm j«, 
pointing probably to p. It is possible that Gill's du in aice, 
auger really means (aau) rather than (oou); for if he had 
vmtten aau^ it would have suggested an approximation to 
(sBseu). 

867. Wallis says : 

^ Au vel awy rect^ pronunciatum, sonum exhiberet compositum 
ex Anglorum d brevi et to, [eeu]. Sed a plerisque nunc dierum 
effertur simpUciter ut Germanorum d pingue [oo] ; sono nempe 
literee d dilatato, et sono litteree w prorsus suppresso. Eodem nempe 
Bono efferunt dU omnes, awl subula ; edU voco, cavX, cowl, omentum, 
vel etiam tiara muliebris.' 

868. Cooper says : 

^A in can^ east, cum u coalescens . . . nunquam occurrit in 
nostr& lingu&. Lofice basta, lancet scalprum chirurgicum, k lanceola ; 
la/nch navem solvere k G. lancer, Jaculari, Ganch in sudes acutae 
prsBcipitem dare, hunt k G. homier frequento ; hanch k G. hanche 
femur; OarU, macer quasi want ab A. S. wana carens, ganCUit cbi- 
rotheca ferrea, landresa k lavando, nuUo mode scribi debent cum u ; 
contr^ enim suadent sonus et derivatio; falsb itaque seribuntur 
launce &c, Queedam vocabula k latinis pnecipue derivata scribimus 
'per au pronunciamus prout au vel a [oo] audacious audax ; maunder 



MODEEN ENGLISH : DIPHTHONGS. 247 

mnnnurare ; k G. mavdire maledioere in loss, lost con- 

juBctoB cum u semper scribimus per au, ut audible aadibilis, 
audi&nce audientia; audtt-or't/ auditorium, aufftnetU augeo, augwry 
augurium, august augustus, auricular auriculans, austerity austeritas, 
authentick authenticus, autliority authoritas, caviious cautus, fravdu" 
lent dolosufl, laudable laudabilis, laurel laurus, plausible plausibilis, 
negligenter loquentes pronuuciant prout a ; in ceteris vocibus ou & 
aw semper prout a pronunciamus/ 

Cp*B occasional (ou) reminds us of Gill's (oou)^ both being 
probably the intermediate stage between (au) and {00). W/s 
(aeu) seems to be a purely theoretical pronunciation. 

869. In thMn the monophthong became universal. The 
sound is now narrow— jf— the earUer sound being probably j*. 
Before U it is now shortened to j, as in salty malt = fMn (sault, 
mault). 

860. In some combinations au dropt its u^ and was treated 
like a, as in (laaf, lsLSi£t9T)=langA(ter) through sMn (Isef, Isesef). 
So also where I is dropt after parasite u in (haaf, haav) = A(df, 
halve through (haef) etc. In (aamz, aans9r)=:a/^ answer ^ fMn 
(aulmz, aunser) — ^which owe their au to the analogy of the 
Anglo-French au^'Et a before nasals in aunt^ daunt etc — au 
seems to have passed straight into (aa) after the older aa had 
become (sese). (aant)=a»^ * formica' also points to a fMn 
(aunt) formed on the analogy, of the foreign aunt ^ amita.' 

861. In MnE orthography ME ^u is always written eto, as 
in hew vb,/w, lewd. So also in strew =,0^ str^ian, which 
probably had fu in ME. ME eu (iu) is also written ew in some 
words, such as new, knew^ steward, but in others it is written 
Uy ue, as in hue sb, truey truth, tuesday. In fMn, u(e) = the 
close ME iu is often written in words which now have only 
ewy thus Td has slue, drue=slew, drew. Conversely, Pg writes 
trewe=itrue. This confusion between dose ew and u is the 
result of the IMF, change of final Fr u fe into eu (691), the 
confusion between the traditional spelling vertu{e) and the 
phonetic vertew leading to a similar fluctuation between trewe 
and true, the latter prevailing. The distinction between 



248 HISTOET OF ENGLISH SOUNDS. 

dose and open ew is further shown in Td's constant spelling 
feawe^few (ME fitoe)^ which, at the same time, shows that 
the first element of the open ew — ^and therefore probably of 
the close etc as well — was long. 

862. In HVg and Sb close etc is transcribed uw. Thus 
HVg has truwth, Dnuws *Jews*, where the'* is only in- 
flectional, = ME JeweSy Jue* (OFr Jui^). Sb has truw, vertuw. 
They also transcribe u in words of French origin with ««?, 
not only finally but also before a cons. Thus HYg has uws 
vb, ffruwt^ Sb has dutok *duke', tre^uwr 'treasure.' It is 
evident that the ME diphthonging of final u had now been 
extended to every u, Welsh uv> is li, and its use in these 
transcriptions must be taken as proof of a diphthongal 
pronunciation in the E. words cited above. If such a word 
as duke had preserved its f* as a monophthong, HYg and Sb 
would have written it simply ^duk^ parallel to hun (809) ; and 
there can be no doubt of the diphthongal character of the 
final ue in true etc, for it was already ME. It would be 
possible to explain uvo as an attempt to indicate a sound 
between I and I, which f might be regarded as, but this is 
against the general principles of the Welsh transcriptions, 
which simply identify each E. sound with the nearest Welsh one. 
The u in this uw cannot well represent any other sound than 
f in E. : we must, therefore, assume that in fMn ME eu and u 
were diphthonged into &. The most probable explanation is 
that eu became (iiu) by the regular change of e into (ii), and 
that the (ii) was rounded by the following «, the resulting 
(yyu) or (yu) afterwards supplanting the non-final as well as 
the final u, 

868. Unfortunately neither HYg nor Sb give a single 
example of open ew. We must, therefore, turn to Palsgrave. 
He says : 

* Ev in the irenche tong hath two djuerse Boundjnges, for sometyme 
they sound hym lyke as we do in our tonge, in these wordes a dewe, 
a shrewe, a fewe, and somtyme like as we do in these wordes, (retoe, 
glewe^ rewe, a meioe. The soandyng of w, whiche is most general in 
the frenche tong, is snche as I haue shewed hy example in these 
wordes, a dewe, a shrewe, a fewe, that is to saye, lyke as the Italians 



MODESN ENGLISH: DIPHTHONGS. 249 

sound ev, or they with vs, that pronounce the latino tonge aryght, as 
&VTevx, vrwXy Jmo, dtw* 

' Uy in the frenche tong, wheresoeuer he is a vowel by hymselfe, shall 
be Bownded like as we sownde eto in these wordes in our tong, rewe 
an herbe, a kmw for a hauke, a clew of threde, and such lyke restyng 
apon the pronounsyng of hym : as for these wordes plus^ nv2, fus^ 
H8^, hiUmble, uertH, they sound jtUvus, nevuly/evua, wumt^ hevwmhU, 
uerievu^ and so in all other wordes, where v is a Yowel by hymselfe 
alone ; so that in the soundynge of this vowel, they differe both from 
the Latin tong and from vs/ 

We ajre here told that the open ew in dew, shrew, few = OE 
diaw, screawa, feawe, ME dfw etc was pronounced as the 
Italian (eu), while the close ew in true etc has the 
Fr sound f«. The first statement supplements Sb in the 
manner we would expect, the second differs from him in 
making long u a monophthongic f ♦. 

864. The other fMn authorities distinctly analyse open 
eiw into (e) + (u). Smith identifies it with the Greek diph- 
thong en, giving as examples :• (feu) * pauci/ (deu) 'ros,* (meu) 
'vox catorum,' (Jeu) 'monstrare,' (sti*eu) 'spargere.* Again: 
'171; sonamus apertius, vt illud Oallicum heau, quod multi 
Angli lev, : sonum etiam felium quidam mjew, alii m/eau, quasi 
/icv, \Lr\v exprimunt.' BU writes A^='hew' with a comma 
under the u to indicate that it has the sound (u). GUI 
lengthens the first element : ' E. saepiils prsecedit u, vt, in 
(eeu) EAWE ovicula, (feeu) fewe pauci, (seeuer) seweb 
dapifer." 

865. These same authorities agi*ee in considering dose ew 
and long f^ to be a simple f «. Cheke says : 

' Cum duke tuke lute rebuke hvK tvk Xvt ptpvK dicimus, Gnecum 
V sonaremus,' of which he says * simplex est, nihil admixtum, nihil 
adjunctum habet.' 

Smith says : 

^Y vel V Grsecum aut Gallicum, quod per se apud nos taxum 
arborem significat, taxus v.' The following are lus examples : ' (snyy) 
ningebat, (slyy) occidit, (tryy) verum, (tyyn) tonus, (kyy) q. litera» 
(ryy) ruta, (myy) cavea in qua tenentur aocipitres, (nyy) novum ; 
(tyyli) valetudinarius, (dyyk) dux, (myyl) mula, (flyyt) tibia 



250 HISTORY OF ENGLISH SOUNDS. 

Oermanorum, (dyy) debitum, (Ijyt) testndo, (bryy) cenuBiam facere, 
(myylet) mulus, (blyy) cfieruleum, (akkyyz) accusare.' 

^Quod genus pronunciationis nos k Gallis accepisse argoit, qu5d 
rarius quidem nos Angli in pronuntiando hac utimor litera. Scoti 
autem qui Grallica lingua suam yeterem quasi obliterarant, et qui 
trans Trentam fluvium habitant, vicinioresque sunt Scotis, frequen- 
tissimd, adeo vt quod nos per Y Eomanum sonamus (u), illi libenter 
proferunt per v Grsecum aut Gallicum (yy) j nam et hie sonus tarn 
Gallis est peculiaris, ut omnia fere Romane scripta per u et v pro- 
ferunt, vt pro Dominus (Dominyys) et lesvs (Jesjys), intantum 
vt qu8B brevia sint natura, vt iUud macrum v exprimant melius, 
sua pronunciatione longa faciunt. Hunc sonum Anglosaxones, de 
quibus postea mentionem faciemus, per 7 exprimebant, ut verus 
Anglosaxonice tiny. Angli (huur) meretrix, (kuuk) coquus, (guud) 
bonum, (bluud) sanguis, (huud) cucullus, (fluud) fluvius, (buuk) 
liber, (tuuk) cepit; ScoH (hyyr, kyyk, gyyd, blyyd, hyyd, flyyd, 
byyk, tyyk)/ * And again, ' O rotundo ore et robustiiis quam 
priores effertur, u angustiore, caetera similis r^ o. Sed v com- 
pressis propemodum labris, multb exilius tenuiusque resonat qukm 
o aut u (boot) scapha, (buut) ocrea, (bjyt) Scotic& pronunciatione, 
ocrea,' 

The Scotch u is now {^, but it may have been closer in Smith's 
time. Note that Sm. gives (yy) as the pronunciation of yew^ 
where we should expect (jyy). 

866. Hart calls u long a diphthong, and writes it iti, but he 
calls Fr u, with which he identifies his E. iu, a diphthong 
also, and it is dear from his description that in his iu the 
front and lip action was simultaneous, giving f«, so that with 
him ^diphthong' means simply ^compound': 

' Now to come to the t«. I sayde the French, Spanish, & Brutes 
[Welsh], I maye adde the Scottish, doe abuse it with vs in sounde and 
for consonant, except the Brutes as is sayd : the French doe neuer 
sound it right, but vsurpe ou, for it, the Spanyard doth often vse it 
right as we doe, but often also abuse it with vs ; the French and 
the Scottish in the sounde of a Diphthong : which keeping the 
vowels in their due soimds, commeth of i & u, (or verie neare it) is 
made and put togither vnder one breath, confounding the soundes of 
i, & u, togither : which you may perceyue in shaping thereof, if you 
take away the inner part of the tongue, from the upper teeth or 



MODEBN ENGLISH: DIPHTHONGS. 25 1 

Qummes, then sliall you sound the a right, or in sounding the 
French and Scottish u, holding still your tongue to the ypper teeth 
or gums, & opening your Uppes somewhat, you shall perceyue the 
right Bounde of ».' 

867. Baret says : 

' And as for the sound of Y yowel ^ whether it be to be sounded 
more sharply as in spelling blue or more grosly like 00, as we 
soun^ Booke, it were long here to discusse. Some therefore think 
that this sharpe Scottish Y is rather a diphthong than a yowell, 
being compounded of our English e and u, as indeed we may partly 
perceyue in pronouncing it, our tongue at the beginning lying flat 
in our mouth, and at the ende rising up with the lips also there- 
withall somewhat more drawen togither.' 

ThiB statement that long u begins with a low-mixed yowel 
— for such would be the result of the tongue lying flat in the 
mouth — cannot be accepted. The most probable interpreta- 
tion is that of Mr. Ellis's, yiz. that Baret was thinking of the 
nentraJ position of the tongue before beginning to utter any 
sound. The whole passage giyes the impression that Baret pro- 
nounced ft, but was trying to conyince himself on theoretical 
grounds similar to those of Hart that it was a true diphthong. 

868. Bll says that long u has a ' sharp ' sound, which he 
identifies with Fr u. 

869. Erondell (1605) says : 

' V Is sounded without any help of the tongue but ioyning of the 
lips as if you would whistle, say u, which u, maketh a siUable by it 
selfe, as vnir, tmiquement as if it were written tMieer, pronounce then 
musique, punir, wbvenir not after the English pronounciation, not as 
if it were written muesiqus, puenir, suevenir, but rather as the u in 
this word, murtherer, not making the u too long.' 

This statement, obscure as it is, seems to agree with Sb's. 
He finds f in E. only as a short yowel, and although his ue 
for the long E. t» is unintelligible, it certainly points to a 
diphthongic pronunciation different from (iu), which he would 
haye expressed by iou^ as Holiband does (870), and which 
therefore may haye been (yu). 

870. Holiband (1609) distinctly describes the (iu)-80und : 

' Printed eoiuofian/. 



\ 



252 HISTORY OF ENGLISH SOUNDS. 

' Where you" must take paine to pronounce onr v, otherwiae then in 
English : for we do thinke that when Englishmen do profer, v, they 
B&j, you : and for, q, we suppose they say, kiou : but we sound, y, 
without any helpe of the tongue, ioyning the lips as if you would 
whistle ; and after the manner that the Scots do sound Gud/ 

871. Cotgrave in 161 1 says: 

' V is sounded as if you whistle it out, as in the word a ItOe' 
Gill is not very definite, but he gives no hint of a diph* 
tbongal pronunciation of u long, calls it vyjftKdv, and his 
description does not contradict that of the others : 

' V, est tenuis, aut crassa : tenuis v, est in Verbo tu yz ysb utor ; 
crassa hreuis estJi,vtin pronomine us tios ; aut longa H: vt in V0r&o 
tu tiz OOSE scaturio, aut sensum exeo mori aqua vi expresses.' 

872. Butler says : 

'/ and u short have a manifest difference from the same long; 
as in ride rid, rude rud, dine din, du/ne dtm, tine tin, tu/ne twn, ; for 
as t short hath the sound of ee short ; so has u short the sound of 00 
short. , , . E and t short with to have the yery sound of u long : 
as in hiw, kneew, Prue appeareth. But because u is the more simple 
and ready way ; and therefore is this sound rather to be expressed 
by it. . . . But why are some of these written with the diphthong ew 1 
whose sound is manifestly different, as in dew^ ewe, few, hew^ chew, 
rew, sew, strew, skew, shrew, pewter.' 

This statement is so ambiguous that we cannot teU whether 
he means that u long was pronounced (iu) or that tV, eeto 
were pronounced (yy). As we shall see, the (iu)-sound was 
fully developed in the next period. All we gather with 
certainty from this statement is that open eia in dew etc was 
distinct from close ew and long u, 

878. In sMn we still find Wallis insisting on the (yy) 
sound : 

' Ibidem etiam,' that is, in lahiis, ' sed Minori adhuc apertura,' than 
(uu), ' formatur ii exile ; Anglis simul et Gallis notissimum. Hoc 
Bono Angli suum u longum ubique proferunt (nonnunquam etiam eu 
et ew quse tamen rectius pronunciantur retento etiam sono e masculi : 
Ut muse, musa, tune, modulatio, lute, barbitum ; dure, dure ; nrnte, 
mutus; new, novus; brew, misceo (cereyisiam coquo); knew, novi; 
view, aspicio ; lieu, vice, etc. Hunc sonum extranei fere assequentur, 



MODBEN ENGLISH : DIPHTHONGS. 253 

b! diphtliongam iu conentur pronanciare ; nempe l exile littersd u 
yel w prseponenteSy (ut in Hispanoram cmdad civitas,) non tamen 
idem est omninb sonus, quamyis ad ilium proximo accedat ; est enim 
iu sonuB compoBituB, at Angloram et Gallomm ii Bonus simplex. 
Cambro-Britanni hunc fere Bonum utcunque per tto, yw, vno descri* 
bunt, ut in lUw color; Uyw gubemaculum nayis ; Duw Deus, aliisque 
innumeris.' 

^U longum effertur ut Gallorum il exile« Ut in like barbitum, 
miUe mutuB, wdse musa, c(Mre cura, etc. Sono nempe quasi composito 
ex i et w,' 

Here Wallis, while pointing out the resemblance between 
Spanish 1%, Welsh iw J^, yio^ uw & on the one hand and Fr ft 
on the other, states expressly that Spanish iu is a diphthong, 
Fr (yy) a simple sound, and with this latter he identifies the 
E. u long and eu^ ew in some cases (meaning, of course, close 
ew). In contradiction to Sb he allows only resemblance to, 
not identity with the Welsh uw^ which he eyidently heard 
as .b — its present sound in South Wales. 

874. To Wallis's contemporary Wilkins, on the contrary, 
the Fr i» is entirely foreign ; he says : 

' As for the u GaUicum or ufhisUing u, though it cannot be denied 
to be a distinct simple yowel ; yet it is of so laborious and difficult 
pronunciation to aU those Nations amongst wbom it is not used, (as 
to the English) especially in the distinction of long and short, and 
framing of Diphthongs, that though I haye enumerated it with the 
rest, and shall make proyision for the expression of it, yet shall I 
make less use of it, than of the others ; and for that reason, not pro- 
ceed to any further explication of it/ 

Accordingly, he transliterates communion by (kommiuunion). 

876. Holder describes (yy) very accurately (795), and says 
that it naturally follows (se) and (e) in diphthongs Does this 
mean that he pronounced eu (open as well as close?) and u 
long as (iy) ? The example he gives is the Lt eu^e. 

876. Cooper says : 

* JB in ioiU, toeoZ cum u coalescens nobis familiarissimus est, 
quem yocamus u longum ; ut fuiwral funus, huge inus [sic] ; juice 
BUCCUB, BcribimuB per ew; ut cTiew mastico, knew cognoyi; aliisque 
temporibuB yerborum prssteritis; quando syllabam finalem claudit, 



254 HISTOET OP ENGLISH SOUNDS. 

additur e, true venis ; rarb per eu^ rheum rheuma ; sic semper pro- 
nunciamuB eu latinum, & eu Grsecum : et Qalli plerumque iUorum te, 
quandoque autem subtiliiis quasi sonus esset simplex, sed hs&c difficilis 
& GaJUs propria.' 

Cp here compares and distinguishes (m) and (yy) very much 
as W. does, but only admits the diphthong in E.^ agreeing 
with Wilkins in finding (yy) a difficult sound. 

877. Miege hears the E. u long as the Fr «, which is prob- 
ably an inaccuracy of ear or description. 

878. We must now return to the open ew, Wallis says : 

* Eu^ eWf eau eonantur per ^ clarum et w. TJt in neuter neutralis, 
few pauci, beatUy pulchritude. Quidam tamen paulo acutius effemnt 
acsi scriberenter, niewter, J^t hiewty^ vel niwter, fiw, bitoty ; prse* 
sertim in vocibus new novus, knew sciebam, stiew ningebat. At pnor 
pronunciatio rectior est.' 

We learn from this passage that the old (eu)=ME ^ was 
beginning to die out,/<?K7 being generally pronounced with the 
first element * sharper* than (e), which W. expresses by writing 
jfiefo, meaning, if not (fiu), at any rate something practically 
identical with it. But he gives as an alternative notation /«e^, 
adding that this e'K^-pronunciation is especially frequent in 
new and some other (probably all) words with ew=ME eto. 
Now under u (873) he includes new in a list of words pro- 
nounced with (yy). Does this imply that/<w also had the 
(yy) sound when not pronounced (feu), or does it mean that 
few had the (iu)-sound, new and « the other close ^w>- words the 
(yy)-sound ? Why then does he not expressly tell us that new 
was pronounced with 'u exile'? Were it not that W. has 
distinguished (iu) from (yy) with such clearness and accuracy 
in treating of u long, we should be obliged to assume that, 
after all, he was incapable of realising the distinction in 
practice, and that he really pronounced (iu) not only in new^ 
but also in muse etc. But when a competent phonetician like 
W. says plainly that his u long is a monophthong identical 
with Fr u] we are bound to believe him, as long as we base our 
conclusions generally on the statements of contemporary pho- 
netic authorities. The most probable solution of the dilemma 
seems to be this. W. himself pronounced (myyz, nyy), but was 



MODERN ENGLISH : DIPHTHONGS. 255 

familiar with the diphthongic (miuz, niu) which he could have 
heard from his contemporary Wilkins, if from no one else. 
This latter pronunciation he has intentionally ignored, while 
unconsciously admitting its existence by identifying the vowel 
of new with the diphthongic (eu) in the modified form of 
(iu). 

879. Wilkins has (eu) in Aeia. Price says that ew keeps its 
sound in few and lewd and some others (most open ew words)) 
but has the sound of iw in blew^ dew and a number of other 
close ew words. Cp has only (m), and this pronunciation be- 
came general in the next cent., so that ME fu^ eu, u were all 
represented by (iu). 

880. In attempting to sum up the results of the preceding 
investigation, the main question that forces itself on us is, was 
the eMn u long (and close eu) a monophthong or a diphthong 1 
We have conclusive evidence of the (iu)-sound in sMn as 
well as late fMn, and strong evidence of the (yu)-80und in 
the fMn period. But there does not seem to be any direct 
connection between these two pronunciations, which are 
separated by a number of authorities who insist on the (yy)- 
sound with such unanimity, and, in several cases, with such 
clearness of description and accuracy of comparison with the 
known sounds of other languages, that we cannot but accept 
their statement. It seems simplest, therefore, to accept these 
facts, which point to the following conclusions. ME (eeu) be- 
came first (iiu), and then by convergence (yyu), which, by 
analogy, supplanted non-final ME (yy). The (u) of (yyu) was 
then absorbed, (yy) being the result, which in sMn was diph- 
thonged into (iu). Another hypothesis is, that (yyu) was the 
only sound in fMn, which, differing so slightly from (yy), was 
generally identified with it, the first element being afterwards 
unrounded, giving (iiu, iu). The last hypothesis is, that the 
normal fMn pronunciation was (iu), of which Sb's uw is ek 
dialectal variation. If we interpret our authorities as literally 
as we can, the first hypothesis is the most probable ; but if 
we attempt to harmonise their contradictory statements, the 
second hypothesis gives a satisfactory explanation of their 
occasional identification of their (yu) with Fr (yy), for even a 



256 HISTORY OF ENGLISH SOUNDS. 

trained phonetician might have some difficulty in distinguish-* 
ing these sounds. 

881. In thMn (iu) shifted the stress on to the second 
element, giving (juu). Lediard expresses the sound of initial 
u in German letters by JuA in juhnion, generally writing iuh 
non-initially. He gives the rule that «» is a long Gm u or uh 
after the forward conss. d^ /, n^ r, t^ thus carrying the dropping 
of the (j) further than in the present E. He then makes a 
remark which is thus abbreviated by Mr. Ellis : 

' According to Mr. Brightland and others, the English express the 
sound of French u by their long u, and sometimes by eu and ew, 
I cannot agree with this opinioD, for although the English perhaps do 
not give the full sound of German u to their long u after (2, 2, n, r, t^ 
yet their sound certainly approaches to this more closely than to the 
French tc, which has induced me to give the German u as its sound, 
contrary to the opinion of some writers. After other consonants 
English long u is iu^ and has nothing in common with French u.' 

We are here told that the (j)-curtailed u in rtide etc, 
though nearer it than ft, is not identical with the former. 
This remark points to the mixed It, due probably to the 
influence of the lost ( j). 



ou 

882. In MnE, as in IME, ou^ ow is written for the ME diph- 
thongs ou, gu and ou — which latter was probably levelled 
under gu in IME — as in grow, know, bow sb=ME growen, kn^en^ 
iowe, OE growan, cndwan, boga. 

The f Mn parasite diphthong in old (842) was not generally 
expressed in writing. The old-diphthongic ou was sharply dis- 
tinguished from the new-diphthongic 0«=ME (uu), as in bma 
vb, i(w^A=lME bowen, bough (721)= OE bugan, bog. The two 
<ni^ are separated in LE also, the old-diphthongic and parasite 
ou being represented by (ou), as in (grou, nou, bou, ould), the 
new-diphthongic ou by au, as in (bau). There must, however, 
have been a time when the two ou^ were very close in sound, 
for ME (uu) passed through the (ou)-stage in fMn (826), 

888. HVg and Sb transcribe old-diphthongic and parasite 



MODERN ENGLISH: DIPHTHONGS. 257 

au sometimes with (m, sometimes with Oy especially when final. 
The following are the chief examples : 

Ou. low^mugire'A 

9U, ou. sowl, sol, owld, howld, sowld, wowld * would'; slo, 
kno, bo ' arcus' S, kro * comix,' tro ' opinor,' bo * arcus' S. 

The dropping of the to in these words, contrasted with its 
invariable retention in aw=:^TilLE u, points to an indistinctness 
of the second element, due to the length of the first. 

884. Smith says : 

'OY dipktJumffvs Grceea, (ou) et mt (oou). Ex (0) breui & (u), 
diphthongum babebant Latini, qu8B si non eadem, vicinissima cert^ 
est ov Graecffi diphthongo, & proximo accedit ad sonum u LatinsB. Ita 
qiiiB Latini per u longum scribebant, Graeci exprimebant per ov. 
quae per u breuem, per v, quasi sonos vicinissimos. At ex (00) longa 
& (a) diphthongus apud nos frequens est, apud Grsecos rara, nisi 
apud lonas : apud Latinos haud scio an fuit vnquam in vsu. 

(ou), (bou) flectere, (boul) sphaera, (kould) poteram, (mou) meta foeni, 

(sou) BUB feemina. 
«»v. (boou) arcuB, (booul) sinum aut scapbiuii], (koould) frigiduB, 

(moou) metere, aut irridere ob distorqueudo, (soou) Beminare, 

aut Buere.' 

And again in his Greek pronunciation he adds : 

'ov ab omnibuB rect^ sonatur, & u facit Latinum quando produ- 
citur, yt aduertit TerentianuB : differt a>v granditate YociB, vt etiam 
rjv ab fv distinguimuB. 
ov. bowj pob, flectere. a hay mow, fiov, foeni congerieB, a gottme, yovv, 

toga. 
c0v. a howj i3a>v, arcuB. to mow, fuk»v, metere, yel ob torquere. gow, 

yavy abeamuB. 
V. V breue Latinum. a huU taurus. u longum vel ov, a howl, /3ovX, 

globus* a>v, a howle )3a»vX, Sinum ligneum, yas in quo lac serua- 

tur, yel ynde ruri bibitur.' 

Here Sm assigns the pronunciation (ou) to Latin u as well 
as to Gk ou. In E. he distinguishes old- from new-diph- 
thongic ou solely by the quantity of the first element, which 
he makes short in the latter, long in the former. Observe 
the distinction between (boul) 'ball* from Fr boule, and (booul) 
'bowl' from 0£ bolla, a confusion between which led to the 

8 



258 HISTORY OF ENGLISH SOUNDS. 

occasional thMn pronunciation of bawl as (haul). BU and Qill 
make the same distinctions as Sm. 

886. In sMn Wallis says : 

' Ou et 0U7 duplicem 6onum obtinent ; alteram clariorem, alteram 
obscuriorem. In quibusdam vocabulis effertur sono dariori per o 
aperium, et w, Ut in s^ul anima^ sould vendebam, venditum, «ndu7 
nix, kn6w Bcio, «$to sero, sao, 6toe debeo, b6fvl poculom, etc^ qao 
etiam sono et 6 simplex nonnanqaam effertur nempe ante ^ at in 
g&ld aurum, scdld rixor, h6ld teneo, cdJd frigidus, did senex, antiqaus, 
etc., et ante 7/ in p6U caput, rdU yoIvo, tdU vectigal, etc. Sed et hsdc 
omnia ab aliis efferantur simpliciter per 6 rotundum acsi scripta 
essent gdle, sdld, snd^ etc. In aliis yocabulis obscuriori sono efferun- 
tur; sono nempe composite ex d vel t^ obscuris, et to. Ut in 
Ttdiise domus, m^use mas, Idwse pediculns, &dtJ globulus, dwr noster, 
dtU ex, dtol bubo, town oppidum, /dtJ immundas,yDu^Z volucris, bdw 
flecto, bditgh ramus, sow sus, etc. At would vellem, should deberem, 
coidd possem, course cursus, eoiurt aula, curia, et pauca forsan alia, 
quamvis (ut proximo pr»cedentia) per du pronunciari debeant, vulgo 
tamen negligentius efferri solent per 00 [uu].' 

886. Cooper says : 

' in fidl, fole cum u conjunctus constituit diphthongum in coulter 
Yomis, four quatuor, tnouid panifico, mucesco, tjpus in quo res 
formatur ; movlter plumas exuere, pouUerer ayicularius, poultry alites 
yillatici, shouider humerus, soul anima ; in cseteris hunc sonum scri- 
bimus per ante U finalem, vel I, quando prsecedit aliam oonsonantem; 
ut bold audax ; quidam hoc modo pronunciant owJ 

' U g^tturalem [b], ante u Qermanicum 00 anglic^ exprimentem 
semper scribimus per ou ; ut out ex ; about circa ; ou tamen aliquando, 
praeter sonum priorem, sonatur ut 00 \ ut I could possem; ut u 
gutturalis, couple copulo ; at a [oo] bought eroptus.' 

As fMn {00) became (00) in sMn, we should expert Smith's 
(boon) etc to naiTow their first element in sMn. Cp ex- 
pressly states that the first element of E. ou was (00) or its 
short, which he identifies with («), by which he probably 
means that it was narrow (o) — }*. W.'s * apertum ' would 
literally mean [o\ but if so, it would be difiicult to understand 
how the dropping (or absorption) of the w could change (snou) 
into (snoo) with }«. It is therefore probable that by open o 



MODERN ENGLISH; CONSONANTS. 259 

W. meant short }, which he hardly recognizes as a distinct 
sound. The epithet 'open' seems to he meant merely to 
exclude * obscure o'=l. 

887. Price and Miege identify £. au with long o, meaning 
(00), which became fixed in thMn, so that no and inaw were 
levelled under (noo), to be diphthonged into (ou) in LE. 



CONSONANTS. 



888. Initial A, which was preserved throughout fMn and 
sMn, began to be dropt everywhere in colloquial speech to- 
wards the end of thMn, but has now been restored in refined 
speech by the infiuence of the spelling, which has introduced 
it into many Fr words where it was originally silent, as in 

889. Already in ME the alternation of such forms as Aii pi 
hie led to the irregular dropping of the A in the uninflected 
At, That these curtailed forms were preserved in eMn is 
shown by such spellings as Aye^Ai^A in Td and the phonetic 
transcription nei^nigA in HVg, enougA pi enow being, on the 
other hand, an example of the faithful retention of the pho- 
netically divergent forms. The retention of the silent gA in 
such words as AigA^ neigA was no doubt partly due to the 
striving after graphical distinctiveness, the spelling Aye^ Aie 
being reserved for the verb to Ai^=' hasten.' 

890. Sb says of the E. gA : 

*• Oh has the same sound as our [Welsh] eh, except that they sound 
gh softly, not in the neck, and we sound eh from the depth of our 
throats and more harshly, and it is disagreeable to the English to 
hear the grating sound of this letter, so Welshmen in the South of 
Wales avoid it as much as possible.' 

The North Welsh cA is a, the South Welsh sound being c 
without any trilL This * harsh,' 'grating' trill was absent 
from the E. gA, according to Sb, who also teUs us that the E. 
sound was not formed ' in the depth of the throat,' which is 
evidently meant to apply to the front o in nigAt etc Whether 

s 2 



26o HISTORY OP ENGLISH SOUNDS. 

the statement is to be applied literally to the gh in laughs 
ME c, is uncertain. If so, it would imply that this latter 
sound had been weakened to the o in wAat^ which is by no 
means improbable. Sb and the HVg transcribe gh by cA, 
never omitting it except where it was already liable to be 
dropped in ME. 

891. The other fMn authorities indicate a very weak sound 
oigh. Smith denotes it by A, saying: 

* Scio tauht^ nikt^fiht & csetera ejusmodi scribi etiam g adjimcta, vfc 
taught^ night, fight, sed sonum illius g quaerant, qoibas ita libet 
scribere, auras profecto mese nunquam in illis yocibos sozdtam rov g 
poterant baurire.' 

Hart agrees, writing lauAt, oAt=zlaugAt, ougAL So also Bll has 
liAt, bowAt=zligAt, hougAt, Gill uses a stem-crossed A to denote 
the sound, and says : 

< X. eh. Grseconim in initio nunquam vsurpamus, in medio, et fine 
Bsepe ; et per gh, male ezprimimuB: posthac sic [crossed A] scribemus : 
vt in (waixt enux) weight bnouoh satis ponderis.' 

892. It seems clear, then, that the regular fMn pronuncia* 
tion of gA reduced it to a mere breath — probably a breath- 
glide modified by the preceding vowel — Tffoo (=TfIr:o), Joo, 
weakenings of earlier lloo, J-oo, 3<:o. But even in this period 
the front gh must have been silent in the pronunciation of the 
majority. Sm gives both (liht) and (leit) = ligAt, and (feit) = 
fgAU Such forms as (leit) can only be explained as diph- 
thongings of an earlier (liit), itself derived from (liht) by 
absorption of the A. If (liht) had really been generally pre- 
served in the beginning of the 17th cent., it could only have 
been contracted into (liit), which would have been preserved 
unchanged in LE, for the earlier (ii) had already become (ei). 
We see, therefore, that the forms (bixt) etc of Gill are really 
half-artificial blendings of the older (liht) and younger (bit). 
There was no doubt a strong — though, of course, hopeless — 
reaction against the dropping of gh, which was natural at a 
period when all the other conss. which are now silent, such as 
the k and w in know and write, were still sounded. The first 
admission of the dropping of ^^ is made by Gill's contemporary 



MODEEN ENGLISH : CONSONANTS. 26 1 

Butler, who uses a crossed^ to denote it, saying ' the Northern 
dialect doth yet rightly sound ' it, implying that it was lost in 
the South. The lip gh must, however, have been preserved 
longer, for not only does it remain to the present day in such 
words as laugh in the form of (f), but the present (oo) in the 
contracted forms nought etc shows that it must have been 
preserved here also till after the narrowing of {00) into (00) — 
that is, till sMn — for otherwise the contraction of (s^ht) into 
(soot) would have resulted in LE (sout). It must, however, be 
noted that the form (soot) actually occurs in sMn by the side 
of (soot), showing that in some pronunciations the gh in the^e 
words must have been dropped early in f Mn. 

898. We can now proceed to the sMn authorities. Wallis, 
after noticing that initial gh is simply (g), adds : 

* alias vero none dierum prorsus omittitur; syllabam tamen pro> 
ducendam innuit. A quibosdam tamen (pnesertim SeptentrioDalibus) 
per molliorem saltern aspirationem h effertur, ut might potestas, lighJt 
lux, night nox, right rectus, f^hi visus, tfvgh singultus, %Jotigh pondero, 
ioeight pondus, t^xigh quamvis, th6ught cogitatio, wr6ught operatus 
est, bright attulit, taught docuit, sought queesivit, fraught refertus, 
nougld nihil, naught malus, &c. In paucis vocabulis effertar plerum- 
que per^; nempe C(yajgh tussis, Prdtigh alveolus, tough tenax, rough 
asper, laugh rideo proferuntur cqffl trojffl tuff, ''^ffy l^ff* Inough 
(singulare) sat multam, sonatur inuff] at xnou>gh (plarale) sat multa, 
sonatur enovo! 

894. Here Smith's (riht) etc appears only as a Northern 
provincialism. 

896. Wilkins, after saying that gh might have been (j) 
adds: 

' This kind of sound is now by disuse lost among us/ 
Price, however, in the same year, says : 

<Gh sounds now like h in Almighty, aUhough' etc., 

adding in the margin : 

' Bat the Ancients did, as the Welch & Scots do still, pronounce gh 
tborow the throat.' 

He notes that gh sounds as (f) in cough, laughter, enough, rough. 
Cooper says : 



262 HISTORY OF ENGLISH SOUNDS. 

'hodi^ apud nos desaevit pronunciatio gh, retinetur tamen in 
scriptura,' 

but he makes it (f) in cough^ laugh^ roughs tough^ trough^ and 
makes Wallis*s distinction between enough and enow. Miege 
says also that "gh is generally mute, but is (f) in lattgh^ draughty 
roughs tough, enough (not distinguishing enow), but adds : 

* S^ghy un Soupir, et le Verbe to Sigh soupirer, ont un son particulier 
qui approche fort de celui du ih en Anglois/ 

Jones ( 1 70 1 ) extends both the (f) and the (th) list. According 
to him (f) is heard regularly in draught, draughts, laugh, cough, 
enojigh, hough, rough, lough, trough ; and he adds : 

' Some also sound datighter, bought, nought, taught, dcCj as with an 
f, Baying daufier, hoft, &c/ 

And he states that gh, ght are th 

' in eigh, sounded sith ; in drought, height sounded drouth, heith,^ 

but in other parts of his book he also admits the sounds (soi, 
droot, heet). 

896. It will now be desirable to consider the changes of gh 
in connection with the preceding vowel. The following are 
the ME combinations with front gh and their Mn equivalents : 

i(h): highj(hAi), nigh (nai), thigh (fai). 

iht: right (rait) etc through (riit). whit (whit) is an 
anomalous weak form of wight (wait). 

ei(h) : neigh (nei), neighbour (neiber). 

eiht : eight (eit), weight (weit). height (hait) owes its vowel 
to the infl. of high. 

897. The combination back vowel + lip-^A inserts an u 
before the gh, which, however, does not seem necessarily to 
form a full diphthong with the preceding vowel, being some- 
times omitted, as in Td's doghter by the side of doughter, wroght, 
and ochtz^ ought in HVg. Such forms as (soot) sought, (laaf) 
laugh point to an ^^-less pronunciation in fMn, while such as 
(toot) taught postulate a full diphthong. It will be observed 
that final lip gh is regularly preserved in LE as (f), except 
in weak words, such as though, and, of course, where analogy 
has been at work, as in drew: 



MODERN ENGLISH: CONSONANTS. 263 

uh: through (fruu) from Jiruh = OE J)urh is a weak form. 
The strong form would be (^J^ref), as in rtrngh, 

auh : laugh (laaf). 

auht : laughter (laafter), draughty draft (draafb). (n)aught (ot, 
not), slaughter (sloter), taught (tot), daughter (dotor). 

nht : doughty (dauti), drought (draut) through (duuti) etc. 

nh = OE uh : rough (raf). 

uh=OE oh: tough (ta^, enough (inaf). slough (BltkXi)^ plough 
(plan), bough (bau). slew^ drew owe their (uu) = earlier (iu) to 
the analogy of the old ^u^-preterites knew^ crew etc. 

o(u)h : cough (kof), trough (trof). though (ion) is a weak form 
of the obsolete vg ()^f). 

o(u)ht : ought (ot), thought (]70t) etc. 

5(a)h : hough (hok) earlier (hof). (^haS) would be the regular 
form (cp iih = OE oh above). 

9(u)h : dough (dou). 

J 

898. There was evidently a dislike to the combination of 
(j) with its cognate vowel (i), which led to the substitution of 
yes, yet foryis^yity which seem originally to have been the more 
usual fMn forms. Whether Hart's (iild) really means that he 
dropt (j) before (ii) is doubtful, as it may be simply a result 
of his theoreticaDy identifying (j) with (i), as he does. But 
cp w ^920). 



r 
890. Sb says : 

' R is of the same nature in the two languages, except that r is 
never doubled or aspirated at the beginning of words as in Greek and 
Welsh.' 

This identifies E. r with the Welsh r cdi, and excludes the 
aspirate, now written rh ot^. 

900. Ben Jonson says : 

'R is the Dogs letter, and hurreth in the sound; the tongue 
striking the inner palate, with a trembling about the teeth. It is 



264 HISTORY OP ENGLISH SOUNDS. 

sounded firme in the beginning of the words, and more liquid in the 
middle, and ends : as in rarer, viper, and so in the Latino/ 

If taken literally — and there seems no reason why it should 
not — this means that r was a point trill cdi initially, and was 
untrilled — = the present <D — before a vowel. 

901. Cooper, however, says that final r is trilled : 

'Verba Anglicana & latina derivativa quae in origine sciibuntur 
cum er scribimus item er, pronunciamus autem ur [vr], non quia sic 
proferri debet, sed quia propter literse r vibrationem yix aliter efferri 
potest ; ut adder coluber, prefer prsefero, slender tenuis.' * r sonatur 
post in apron gremiale, citron citreum, environ circundo, gridiron 
craticula, iron ferrum, saffron crocus ; quasi scriberentur apwm^ &c' 

But here the mention of the vibration seems to be nothing 
but a part of the traditional definition of r. It is remarkable 
how people cling even now to the idea that the E. r is trilled, 
probably confounding trilling with the voice- vibration in the 
glottis. Walker even imagines a tiill of the root of the 
tongue in one of his pronunciations of r. Lediard (1725) 
says of E. r that it agrees entirely with the Gm r, which at 
that time was no doubt (01 all over North Germany, as it still 
is in the remoter districts, having been supplanted by the 
back e in the towns. 

902. In LE r is (D before a vowel, being always dropt before 
a pause or another cons., leaving only an (8)-glide behind it, 
which is absorbed by a preceding obscure vowel. We. have 
now to trace the development of this voice-glide. 

908. Sb transcribes E. weak er by ^r, yr, tV, r, as in kwaHer^ 
papyr^ tsintsir *gynger,' thwndr^ vmdr *wondre,* which points to 
an indistinct (ar). Bll has special signs for syllabic l^m^nia 
fable etc, but none for syllabic r, which shows that Sb's thvmdr 
really means ( J^frnder). 

904. We have now to consider the influence of r on pre- 
ceding vowels, which has played so important a part in the 
development of the LE vowel-system. The change of e into a 
before r, as in far, reaches back, as we have seen (789), into 
the ME period. The first traces of the specifically Mn 
changes are found in sMn. Wallis tells us (793) that e before 
r, as in vertue^ had the sound of Fr ^e feminine,' which we 



MODERN ENGLISH: CONSONANTS. 265 

have identified as I or X- He expressly distinguishes er from 
wr in ium^ where the u kept its regular sMn sound, but still, 
her^ turn ff^ci), dlCDTf had now approximated their vowels con- 
siderably, and the b^inning was made of that levelling of 
vowels which has now been carried out in such words as her^ 
fuT^fir (hear, foor, fear). Cp, in the passage quoted just above 
(901), seems to identify the sound of er completely with that 
of UT. He also gives the same sound to the ir in hird etc. 

906. The following examples will show how the ME vowels 
have changed before r in MnE and LE. Observe that (r) 
followed by a vowel in the same word has no effect on a 
preceding short vowel ; thus (nserou) keeps the regular short 
(sb) of (maen). 

ar: narrow (ncerou). far^ hard (faar, faa; haad). quarrel 
(kworol). noar^ warm (wor, woe ; wom). 

ir: stirrup (stiT9^), «^}r,^r«^ (steer, stoe; feest). 

er: herring (herii)). her^ were (heer, hee; weer, wee), herd^ 
heardy learn (heed, leen). As er was regularly lengthened to 
^r in ME in strest syllables, except when followed by certain 
conss.^ final er could become (eer) only in weak forms such as 
her and were, which latter is a shortened ME wfre. The (ee) in 
heard, learn etc points to eMn (herd) etc by the side of (heerd). 
tarrjf (tseri). far, dark (faar, faa ; daak.) 

ur : furrow (farou). spur, further, worth (speer, spee ; feeder, 
weej?). word (weed) points to f Mn (wurd) with shortened (uu) 

:=MEJ. 

or : sorrow (sorou). for, north (for, foe ; no)?). Final -or only 
in weak syllables (cp er). 

fc: care (kter). Cp name (neim). 

It : fire (faier). 

er : deer (dier),/eflr (fier), here (hier). 

fr: ear (ier), tear sb (tier), heard (bied). bear (bter), tear 
(Uer) vb. 

Or : sour (sauer). 

UP : cure (kjuer). h^re (luer). 

5p : moor (muer). jloor (flor). 

9r : more (mor). hoarse (hos). 

air : fair (fter). 



266 HISTOEY OF ENGLISH SOUNDS. 

eir : tAeir (Scor). 

fix : stairs (stsaz). 

OUT : four (for), fourth (foj?). 



906. In eMn I must have had the same deep pitch as at 
present, as shown by its development of a parasite u between 
it and a preceding back vowel (784, 84a). 

807. (1) is regularly dropt between its parasite (u) and a 
following cons, in the fMn combinations aulf aulv, aulm, auli, 
oulk, as in ialf^ halvCy calm; walk; folk (haaf, haav, kaam ; 
wok; fouk). Also in should^ would^ could (Jud, wud, kud). 
This loss of I began in flfn. In the last three words (the last 
of which,^ME cupe, cude^ owed its / to the analogy of the 
other two) the I seems at first to have been dropped only in 
the weak forms. 

808. Sb notes the provincial dropping of ^ in bot^^ bto, caw^ 
as he writes them,=do&2, bully call. There are traces of this in 
the literary language, for we can only explain Td's curious 
spelling rayneboll= rainbow on the supposition that he pro- 
nounced bowl and bow ' arcus ' alike as (bou). 



h Bf I, f- 
909. In some words Mn£ final (p) corresponds anomalously 
to medial ME M =(*$), as in pit A, beneath^ both, earthy fourth etc. 
Final (z)=ME (s) in arosCy chose is due to the infl. of the inf. 
rise, chose. The (z) of toise a<^. must also be due to some 
analogy — either of the inflected ME wise or of wisdom, 

810. /=ME V in beliefs sherrif and a few others. 

811. The present distinction between initial (J') and (*$) is 
fully confirmed for fMn by HVg, which writes dde, ddetiy ddat^ 
ddei^zthe, then, that, they with the Welsh Ai=('8), writing 
Welsh th=z(]f) in other words. 

812. In HVg with and of are written wyth and 0^= strong 
(wi)?, of). Td writes both of and off for the prp. The other 
authorities give (f>) in with. Hart, however, has (ui'S). There 



MODERN ENGLISH: CONSONANTS. 267 

can, of course, be no doubt that (wi]?, wiiS, of, ov) existed side 
by side in fMn as strong and weak forms respectively, (wi]?) 
is now almost extinct, and (of) is entirely restricted to its 
adverbial function. 

Old. The fMn change of (-S) into (d) takes place mainly after 
r, as in murder, burden^ or before r, where 9 and d were confused 
(931), as in rudder, and /, as in twaddle, fiddle. 

014. sk in fMn does not seem to have differed from the 
present sound. Sb says : 

' Sh when coming before a vowel is equivalent to this combination 
Mf, thus SHAPPB stiapp, SHIPS stiip. Sh coming after a vowel ia 
pronounced tM, thus asshb aisa, wasshb waise. And wherever it is 
met with, it hisses like a roused serpent, not unlike the Hebrew 
letter called Bchin, And if you wish further information respecting 
this sound, you should listen to the hissing voice of shellfish when 
they begin to boiL' 

So also HYg writes eiak^ehake, with the variations syts 
* such,' aisA * ask,' sAio * shew.' 

016. i in Welsh is (J) in such words as ceieio, where it has 
developed out of the combination (sj), but this is a very recent 
development. It is possible, however, that even in Sb s time 
e. was palatalized in this combination — s\o>. This was probably 
also the beginning of the LE (J}-sound of # in such a word as 
nation, which Gill writes (naasion). In sMn Wallis recognizes 
(J) in such words, but Wilkins still writes (resinTeksion)= 
resurrection, and Price (1668} only recognizes * hard s' in pamon. 
Cooper (1685) admits ehure, iAnffar,=sure, sugar 'focilitatis 
caus&,' although he stigmatizes the ^A -pronunciation as bar- 
barous. 

016. Miege (1688) writes eh4re, pennchoun in French letters 
for sure, pennon, states that in the termination -ieum, s sounds 
as French p or /, and writes ^ual, traingient, lejeur, Sjer, kSjer, 
crojer for usual, transient, leisure, osier, hosier, crosier. This passage 
contains the first notice of the sound (5), which had previously 
been known only in the combination (d5)=:y and 'soft' g. It 
is not noticed even by Lediard (1725), who seems to pro- 
nounce decision etc with (J). Sheridan (1780) fully recog- 
nizes it. 



268 HISTORY OP ENGLISH SOUNDS. 



817. Gill distinctly recognizes wh as a simple cons. : 

* IT, aspiratum, consona est, quam scribunt per wh et tamen aspi- 
ratio prsecedit. Illse namque voces quee per xoh scribuntur ; possunt 
atque etiam ad exempla msdonim scribi debent per (hw) aut (hu) ; ita 
enim, nihil aliad inde coUigi qaeat, qu^m quod ex ipso toA, intelligimus ; 
vt (wiil) sive (uiil) yteele nassa, (hwiil) sive (huiil) wheels rota. 
Tamen quia nostra experientia docet, (w) et (wh) veras esse simpli- 
cesque consonas, in quarum elatione (u) suggninnit tantum, non clara 
Yocalis auditur; ideo illud (w) ante vocales aut diphthongos ius 
assignatum obtinebit ; at (wh) mala tantum consuetudine valebit in 
(what) quid, (wheSer) uter & similibus.' 

918. Towards the close of thMn (wh) began to be levelled 
tinder (w), and in the present cent, the change was carried out 
universally, even among those who still retained (h) as a mark 
of gentility. Sut of late years it has begun to be restored in 
Southern educated speech, partly by the influence of the 
spelling, partly by that of Scotch and Irish pronunciation, so 
that in another generation it will probably be completely 
restored. It is now pronounced in unstrest words, where it 
was probably weakened into (w) in the period when it was 
a natural sound. 

919. The now silent w before r is preserved not only in 
HVg, as in vyricUt *wright,* but also by all the other fMn 
authorities. Those of the sMn period drop it. Jones, how- 
ever, says * r- may be «t-/ Lediard (1725) says that in wr the 
V) is * little or scarcely heard, as in wracks wrench etc, in which 
I can only find a soft aspiration before rl The development of 
(0) in wrath and of (o) in the vg (rop) =wrap shows clearly that w 
was not simply dropt before r, but that it first rounded it, and 
then was dropt itself as superfluous : s(d, »(d», (D», O). In LE r, 
whether answering to old wr or to simple r, is often rounded, 
especially in emphatic speaking. Perhaps it is to some such 
practice that Jones is alluding in the remark cited above. 

920. Sb's writing wnder^ w for wonder^ woo would seem to be 
the result of Welsh habits, as also Jones's sMn (um8Bn) = 
woinan. But that there was a real dislike to the sequence 



MODERN ENGLISH : CONSONANTS. 269 

(w) + (u) is shown by ooze =0E icos. w has also been dropt 
after in thong and 90^ the last being eME. 

021. The loss of the w of answer is the result of want of 
stress. 



ng 

022. Gill appears to be the first writer who recognizes (i)) 
as a separate element. He says, leaving his notation un- 
altered: 

' N in illis [literis] est quae nihil mutare diximos : at si k^ aut g^ 
sequatur pauliim minuenda est nostra sententia: neque enim (si 
accurate expendas) plan^ ita profertur in thank et think quemad- 
modam pronunciatur in hand roanus, et nion nok£ nullus. 8ed ne 
adeo nasutuli videamor ut nihil vetustate rancidum ferre possimas : 
quia k^ ibi clar^ auditur, nee congruum esse reor quicquam veritati 
propinquum immutare; monuisse tantum volui, sed te invito non 
monuisse tamen. At si g subsequatar vt in thing res et aong canti- 
cum; quia Bonus literee g ibi nullus est, at semivocalis plan^ alia 
quse ab n non minus distat qu^m fn\ liters ng, una erit ex illis 
compositis, quibus fas esse yolui sonum simplicem indicare, ut in sing 
canta, et among inter, hue etiam refer ilia in quibus g^ ab n, ratione 
sequentis liquids quodammodo distrahitur, a spangl nitella, tu intangl 
implicare.' 

This quite agrees with the present usage, which pronounces 
ng finally as (i)), keeping the g before / (spsengl), as also before 
a vowel, as in (haijgor) hunger^ except where the analogy of 
the forms in which (i}) is final have introduced it medially 
also, as in (sii)9r) singer^ {^W^) ^7 the infl. of (sirj). Medial (g) 
is, however, preserved in the comparison of adjectives, as in 
(lorjgar, loi)gist). The fMn usage was no doubt the same. 



028. Initial k before n is written not only in HVg, as in 
knicht * knight/ but also by all the other fMn authorities. 
The sMn Wallis also allows k in know^ knew. Jones also says 
that initial kn * may be sounded kn,* 



270 HISTORY OF ENGLISH SOUNDS. 

024. Cooper says: — 

* Kn Bonatur ut An ; Jenave nebulo . . . qnaai hruzve etc/ 

Lediard (1725) says: 

'K before n at the beginning of a word is only aspirated, and 
spoken as an h; JcrMcJc hnack, knave hnave, knife hneif, knte hnie, 
knot^ know, knucMe, etc. M. Ludwick says that k before n is called 
t ; Arnold and others declare that it is pronounced d. But any one 
experienced in English pronunciation must own, that only a pure 
gentle aspiration is observable, and by no means so hard and un- 
pleasant a sound as must arise from prefixing (£ or ^ to n.' 

This, of course, means that kn did not become the present 
(n) by mere dropping of the A, but the » was unvoiced by the 
off breath-glide of the £, which was then itself dropt as super- 
fluous. 1 was afterwards levelled under the more frequent 7. 
The same change of kn- into 1 has taken place in Mnlcel., 
where knif is pronounced if ♦>. The tn of the Germans was, 
no doubt, only a clumsy way of indicating the voiceless n. 

826. Initial gn does not occur in the fMn authorities, but 
was no doubt (gn), parallel to (kn), Jones in sMn makes it 
simple «. Lediard, however, says : 

' Initial g before n sounds as an aspiration or A, not like a hard g, 
as gnash hnasch not gnasch, gnai hnat not gnat, gnaw hnah not gnah, 
gnomon^ gnostick! 

It is possible that gn- was levelled under the more frequent kn^ 
but a comparison of this statement with that about wr-^ (919)9 
where the term * aspiration' is used without any apparent 
meaning, makes it altogether doubtful. 

826. The old-fashioned fronting of k and g after (aa) in 
(kjaat, gjaadn) carty garden etc, is evidently the result of the 
sMn pronunciation of these words with (aeae). When (aese) 
became (aa) the preceding front-modified ^s and g^ retained 
and exaggerated the front glide on to the (aa). 

t,d; t/,ds 

827. The change of (tj, dj) into (tj, ds) in thMn, as in 
(neitjor, voedsor) nature, verdure, through (nsesetjur) etc is quite 
parallel to that of (sj) into J (915). 



MODEBN ENGLISH : CONSONANTS. 27 1 

The old chyj had abready developed into their present sound 
of (tj, ds) in f Mn, as shown by the insertion of i and d^ which 
is common in Td, as in fetche^ loatche^ knoledging by the side of 
hwwlege. 

028. The voicing of ME ch in knowledge = ME kngtcl^che is 
evidently parallel to the (z) of speechee etc, being due to want 
of stress. In the LE (grinids, wulids) = Greenvnch (OE 
Grenafaic)y Woolwich^ the same change has taken place. We 
may therefore confidently assume an earlier alternation of 
strong (eetj, whitj, sutj) each^ whick^ stick with weak (eeds, wids, 
suds) in whickever etc, and this is confirmed by Lediard's (1725} 
(iids, whids). 

820. In LE ^ preceded by the hisses e and / and followed 
by the vowellikes /, «, m is regularly dropped, as in ({'isl, 
faasn, tjesndt, krismos) thutUy fasten^ ckestnuty chrisimae^ {^^) 
often. It will be observed that in most of the examples the 
vowellike cons, is final, and therefore syllabic ; it is probable 
that the dropping of the i began in such words. In fMn the t 
was still preserved, as shown by Sb's transliteration tkyetl etc, 
but not in sMn, so that Buchanan's preservation of it in thMn 
must be a Scotticism, the t in castle etc being still preserved 
even in refined Scotch pronunciation. 

880. The triple consonant-groups (ItJ, ntj) are lightened in 
the same way by throwing out the ^, as in milck^ benck. So 
also (nd5) becomes (ns), as in einge^ (Ids) ^ ^ ^^^ being kept. 

881. ME d preceded by a vowel and followed by r — ^gene- 
rally with a vowel between — became ("B) in many words in 
£Mn, such ^Afatker^ gatker, togetker, kitker, motker. 

882. The change of t into (p) after r in 9toartk{y) seems to be 
not earlier than thMn. 



p, b 

988. The loss of final b after m occurred already in fMn, 
thus Gill has (lam) = lamb. Td has lambe but domme, domm s= 
dumb, which shows that both pronunciations must have 
existed in the earliest fMn. Such spellings as limb and numb 
ss ME liMy numen seem, indeed, to point to a complete con- 



272 HISTORY OF ENGLISH SOUNDS. 

fusion between final m after a short vowel, and ml, in pro- 
nunciation as well as spelling, larah being pronounced (lam, 
lamb), limb being pronounced (lim, limb) indifferently ; perhaps 
the b was only sounded before a vowel beginning the next 
word. Unstrest b after m was dropt in writing as well as 
pronunciation in oakum^OK dcumba. 

034. b has, on the contrary, been inserted between m and I 
in such words as thimble, bramble. This insertion began in 
IMK 



LIVING ENGLISH SOUNDS. 

936. If we compare the fMn orthography with that of LE, 
we see at once that the latter is distinguished (i) by its entire 
dissociation from the spoken language, and (2) by its fixity. 
The present £. orthography is practically a system of letter- 
groups which are partly arbitrary hieroglyphs, partly im- 
perfectly phonetic representations of the language of the i6th 
cent. 

036. If we compare the sound-systems of the two periods, 
we are struck (i) by the great changes that have taken place 
— changes which have not been in any way retarded by the 
increasing fixity of the orthography — and (a) by the greater 
uniformity of the present pronunciation, which is the result of 
the greater facilities of communication. 

837. In the fMn and sMn periods the influence of spelling 
on pronunciation seems to have been very slight. But as 
standard E. came to be spoken over a continually widening 
area, and as the distinction of polite and vulgar pronunciation 
developed itself more and more, there arose a strong reaction 
against the colloquialisms of the sMn period, and in the thMn 
period many older pronunciations were restored by the influ- 
ence of the written language, schoolmasters and pronouncing- 
dictionaries working hand in hand. Thus, in the 17th cent, 
such a pronunciation as (b8ek9rd)=iac^ar^ was the regular 
one, and our present (baekwed) would have seemed — what it 
no doubt was at first — a pedantic following of the spelling. 
We see the same process in the present pronunciation of 



LIVING ENGLISH: STRESS. 273 

towards as (towodz) which seems likely to supplant the older 
and still commoner (todz). 

838. In the case of words which have become rare and 
obsolete, a refashioning of the pronunciation through mis- 
interpretation of the spelling is inevitable, as in our present 
pronunciation of behove as (bihouv) instead of (bihuwv) with 
the regular (uw)=M£ 0. These influences have not so much 
affected the E. part of our vocabulary as they have the Fr 
and foreign words, where, indeed, the corruption of spelling 
as well as of pronunciation — the latter the consequence of 
the former — have been carried to such an extent as to make 
our present written language almost useless for purposes of 
historical investigation. 



STRESS. 

830. The most characteristic feature of LE is the extreme 
sensibility of its vowels to stress-influence. Most words which 
occur frequently as unstrest members of a sentence develop 
a weak form alongside of the original strong form by modifica- 
tion of the vowel and occasionally by consonant-dropping. 
Thus, in the sentence (-"Saz noubadi "Ssa) ihete is nobody there^ 
("Ss) is the weak form of the strong ({Seo). Here there is a 
distinction of meaning, but in many cases the strong form is 
simply more emphatic than the weak one, as in (:whot *aa -ju 
duwig) what are you doing ? compared with (whot -9 -ju duwig). 
Unstrest vowels all tend to weakening, generally in the 
direction of the mixed vowels, and there are several vowels 
which occur only in unstrest syllables: !▼, Xi Ii I. as in (meni, 
bet9, vtelju, felou >()a)X%)) many^ better^ value^ fellow. The last 
two are weakenings of (uw) and (o, 0) resp. The first two are 
weakenings of a variety of vowels and diphthongs. 

840. The history of MnE and LE stress and stress-influence 
offers great difficulties, because of our defective knowledge of 
the earlier periods. It is certain that many of our present 
weak forms are of sMn, some of fMn origin, while the alter- 
nation of such forms as (hiz) and (iz) can be traced back to 
OE (500). 

T 



274 HISTORY OF ENGLISH SOUNDS. 

841. The history of MnE sentence- stress cannot be attempted 
at present. The most characteristic feature of LE as com- 
pared with OE sentence-stress is its development of level 
stress. In such an attributive group as (big blsek dog), where 
in OE the first element would have had a stronger stress than 
the others, we stress all three words equally. We even 
separate the elements of a traditional compound in the same 
way, if the first word is attributive, as in (ijvnig staa) evening 
«/ar=OE 'afen-sfeorra, and, what is still more remarkable, we 
isolate inseparable prefixes by means of an independent stress, 
if they have a full meaning, as in (-an-kuw)?) uncouth =.OEi 
^un^cup, the Scotch unco (wjko) still preserving the older stress. 
All this, as well as the many delicate gradations of stress 
which in our syntax supply the place of inflection, must be of 
comparatively late origin. 



QUANTITY. 

942. In LE long vowels occur only finally, as in (faa) far^ 
and before voiced conss., as in (haad) hard^ being shortened to 
half-long before voiceless conss., as in (haat) heart. 

843. Final voiceless conss. are short after a long vowel, as 
in (haat), long after a short vowel, as in (hset) hat. If a final 
voice cons, follows a short vowel, as in (bsed) had^ the length 
seems to be generally distributed over both vowel and cons., 
although it is sometimes confined to the vowel. In vgE 
there is a tendency to lengthen the vowel before voiceless 
conss. as well. 

844. These rules apply only to strest syllables, unstrest 
syllables being generally short. A strest short vowel is 
never followed by a short single cons., unless the cons, is 
followed by a short stressless vowel without any pause 
between them, as in (beto) letter. If such a word is drawled 
out, the length is thrown on to the stressless vowel, as in 
(:whot 9 piti) :OfD -^ dIdI*, tthat a pity ! 



LIVING ENGLISH : VOWELS 275 

VOWELS. 

945. The following is the LE vowel-system, weak vowels 
being marked by a preceding ( - ) : 

te u, -u o, -o 

(j)uw ou, -ou ; oi 

(j;u9 0, 09 



a -9 
ai, au 
aa 99 


1. -I 

• • 

is 

r,-r, 
re- 
ft 


e 
ei 

C8 


y -X 





I 1> -5P J, -l 

(<!))h ffc, ji 

046. The first row consists of manqpAlAofiffs, all of which, 
though normally short, occur also long (943), the only monoph- 
thong among the normally long vowels being (99). The 
remaining diphthong-vowels may be classified as divergent 
diphthongs (ai, au, oi), mid diphthongs (ei, ou), Air^A diphthongs 
(ij, uw), murmur diphthongs (i9, E9, u9, 09), and murmur longs 
(aa, o). In these last the murmur is only just audible, while 
in (99) it is completely absorbed. There are also the triph- 
thongs (ai9) etc. All these (9)s are parasite developments 
due to a following r (905). 

947. (a)=(i) ME w, as in (kam) come, {%) «, as in (kratj) 
crutch. (3) 0, as in (avn) oven, (4) J, as in (dan) done. The full 
back ] is heard in the West of E. and in Scotland. In Vg 
this vowel tends to widening and lowering, becoming nearly 
X. American and Irish E. agree in making (a) a sound inter- 
mediate to ] and I — Jw. The StE sound must be the older, 
as being nearer the sMn 1 or 1. 

948. (i) = (i) », as in (liv) live, {%) «, as in (mil) 7nilL (3) 
^, as in (strirj) string, (4) ^, as in (sik) sick. In children (i) 
has been absorbed by an (u)-modified (1), the glide between 
them developing into a full (u) — (tjuldron). In milk the same 
rounded (1) has become syUabic, and the preceding vowel has 
become a glide-vowel — (mjlk), sometimes (mjulk). 

949. (e) = ( I ) ^, as in (west, best) west^ best. (2) 0, as in (meni) 
many, (3) i [i) in (Jerif) sheriff, (4) f, as in (dred) drearl, 
(5) ^, as in {trejn^) friend, (6) «, as in (beri) bury, 

960. eB=(i) a, as in (meen) man, (2) <?, as in (J^raeJ) thresh^ 
thrash, (3) in (strsep) strap. Tends to |) in Vg. 

T 2 



276 HlSTOfiY OF ENGLISH SOUNDS. 

961. u=(i) «, as in ({u\)full, (2) (?, as in (Jud, wud) sAould, 
would. (3) u in (kud) could. (4) J, as in ({ut) foot. 

862. o=:(i) 0, as in (lot) lot. (2) a, as in (wont, solt) icant^ 
salt. (3) g in (hot) ^(?/. 

863. ai=(i) it as in (laif) /j/^<?. (2) t?, as in (braid) bride. 
(3) e^, as in (brait) bright. (4) e in (braiar) iriar. Broadened 
to Hi^T, jf ▼, jIt in Vg, the second element being often obscured — 
j%. Before (1) it is almost completely absorbed in Vg, mile 
being confused with marie. 

964. au=(i) «, as in (haus) house. (2) uh in (dauti) 
doughty. Becomes j^ in Vg. 

866. ij=(i) #, as in (hij, strijt) he^ street. (2) ^, as in (ijst) 
east. (3) /, as in (ijt) ^a^. In Vg the first element is lowered, 
so that the diphthong approximates to -(ei). 

866. ei = (i) i, as in (neim) name. (2) ai, as in (dei) day. 
(3) ei, as in ("Sei) they. (4) ^1, as in (hei) hay. (5) ^, as in 
(klei) c/fl^. In Vg the first element is broadened to \, so that 
this diphthong is confused with St (ai), except when this latter 
is broadened (953). 

857. uw=(i) Oy as in (kuwl) cool. (2) «, as in (stuwp) 
stoop. (3) u in (bruwz) bruise. (4) ^, as in (straw) strew. 
(5) eUy as in (gruw) grew. Becomes If) in Vg. 

858. juw=(i) u, as in (tjuwn) tune. (2) eu, as in (njuwt) 
ne^ot. (3) ^w, as in (njuw) new. (4) p^, as in (Quw) few. 
Becomes <dT3B in Vg initially, T?) non-initially, as in (nuw) new. 

858. ou=(i) p, as in (stoun) stone. (2) ^, as in (koul) (?<w/. 
(3) 0, as in (wouk) woke. (4) ou^ as in (floun) ^<??r«. (5) o{l), 
as in (fouk, boult) /olk^ bolt. (6) o«, as in (flou) flow. (7) 
gUy as in (slou) */o«?. Becomes \\y in affected, 3* in Vg speech. 
Weak (ou), as in narrow, becomes (a) in Vg. 

860. oi=(i) oi, as in (vols) voice. (2) u {i) in (boil) i(>27 sb. 

961. aa=(i) a, as in (graas, haad, haaf) grass, hard, half 

(2) e{r), as in (staar) star. Broadened to j*, j* in Vg. 

862. ee = (i) ir{ur), as in (boatj, baodn) birch, burden. (2) er, as 
in (09J?) ^ar^^. (3) ur, as in (taaf) ^wr^'. (4) or in (w9ad) «?<?rr/. 

963. io = (i) er, as in (stiar) steer. (2) ^r, as in (nier) near. 

(3) ^r, as in (spiar) spear. 

964. se=(i) ^/•(?), as in (hear) hair. (2) p*, as in (car) ere. 



LIVING ENGLISH : CONSONANTS. 277 

(3) Sty as in (swter) %wear. (4) aiV, as in (f£or)yfl/r. (5) eir^ as 
in ("Btar) ^^^ir. (6) ^r, as in (st£Oz) 9tair%, 

965. ue=r^, as in (mudr) moor. 

966. juo=«r, as in (kjuor) cure. 

967. o=:(i) auy as in (dro) e/ra«7. {%) a{l\ as in (fol, wok) 
faUy walk. (3) (ff)a, as in (wotar, worn) watery warm. (4) {w)^r\ 

as in (J?wot) thwart. (5) <?r, as in (hos) hor^e. (6) pr in (hos) 
hoarse. (7) o«r, as in (foti) forty. (8) e>w(i), as in (kof, }>ot) 
rt>i(^A, thought. (9) p in (brod) ire?<M?. 

968. oo=(i) or, as in (foar) for. (a) pr, as in (moor) mare. 
(3) (^, as in (bifoar) before. (4) (wr, as in (fo9r)/(?«r. 

The (e) is dropt when a cons, follows: cp (bifaa) with 
(-bi forit) before it, 

969. The characteristic feature of the LE vowel-system is 
its diphthonging of all the earlier long monophthongs. The 
diphthonging of (ei) and (ou) was first noticed by Smart in 
1836, but it is probably older, as it occurs also in American 
E., which still pronounces !♦, !♦ for our (ij, uw). The broaden- 
ing of (ei, ou) to (ai; au) is not old : it was almost unknown 
thirty years ago, but is now beginning to push its way into 
educated speech. 

CONSONANTS. 

970. The following is the LE cons.-sy8tem : 

h - - - )^, 8, f f , wh 

k - t,tf - p 



— 


J 


r 

1 

d,d5 
n 


«, z,5 


V, w 


g 
9 


— 


- 


b 

m 


— 


— 




v, 8, e 


>, 


a 


— 


0, oe 


— 


D 


— 


— 


— 


— 


~ 


— 


(D 


00 

0), a)e 


w, 5, e 


>, 


Q 


^ 


— 


D 


d 


-. 


T 


— 


F 



278 HISTORY OF ENGLISH SOUNDS. 

971. The L£ consonant-syBiem differs comparatively little 
from the ME. The ME back open c, o, c are wanting, and 
ME cky g have developed into oe, cpe. LE has also developed 
a voiced (J) by the fronting of older (z), as in glazier. 

972. Otherwise the main changes are the loss of initial 
conss. before another vowellike conss., as in LE (n)=ME i«-, 
gnry (r)=ME «v-, and the dropping of (r) when not foDowed 
by a vowel, the last being a specially LE change. 

073. In Vg — as also in most of the LE dialects (but not in 
Scotch, Irish, American and Australasian) — (h) is dropt, being, 
on the other hand, sometimes retained or added before an 
emphatic vowel. In Vg — as also generally in Southern StE — 
(wh) is levelled under (w). Vg always, and StE often, level 
final (e) under (ar), adding an (r) before another vowel as in 
(ai'diar ev) idea of, Vg treats (aa, o) in the same way, as in 
(aar -ai doun nou) ah^ I do not know, Vg changes final weak 
(g) to (n), as in (drorin) drawing. The older Vg (w) for (v) is 
now extinct. 



F1K8T WOBD-LIST. 279 



le 

Mi 

i 



FIKST WOKD-LIST. 

(old-middle-modern.) 

The following list is intended to include the majority 
of the words of 0£ or Scand. origin still in common use. 
The first column gives the OE forms, Scand. words being 
marked %. Words which do not occur at all in 0£, or do not 
occur in the form in which they are here given are marked^. 
Words whose later form-changes are irregular, owing to 
external influences, are marked f. The second column gives 
the ME forms, if possible, those of the Ormulum, which are 
marked f. The thiird column gives the present spelling. 
The fourth column gives the present pronunciation, words 
more or less obsolete in colloquial speech being marked t* 

The notes give the various forms for the four periods— Old, 
Middle, Modem, and Living — each period being sepai'ated by 
a dash. If the first note is preceded by a dash, it applies to 
the Middle period, and so on. When the name of a text etc 
is not preceded by any form, it refers to the heading of its 
period; thus in i FP means that earun is the form in that 
text, while the note on 7 refers to the second (ME) column. 
ME forms which occur in unambiguous rhymes are marked f 
in the notes, or else a specimen rhyme is given, the rhyming 
words being joined by (:). ME forms in ( ) are from some 
other than the chief ms : from Lay.^, AR^, from any ms but 
Ellesm. of Ch. MnE forms in ( ) are transcriptions of the 
phonetic spellings of the phonetic authorities. 

The OE forms are an*anged primarily under their vowels in 
the following order : a (ee, ea), i, e (g, eo), u, y, o, ce ; a, ae, se 
(=non-WS e), e (oe), ea, eo, i, u, y, 6. Scand. ei and oy go 
under e^, Scand. gu under 0. The words are then arranged by 
the cons, which follows the vowel, and lastly by the first 
cons, that precedes the vowel, both in the following order: 
h-, r, hr, 1, hi, )?, s, w, hw, f ; nc, ng, n, m ; c, -h, g, t, d, p, b. 



28o 



HISTORY OF ENGLISH SOUNDS. 



a (869 ea). 



earun vb farm are aar 

VP. arun Du, Ru. — faren, are Hv. arn AllP. fare Aud. fer North, (i^fl. 
ofSoand. era), ar TM. — (aar, ar) G, (»r) 110^ (eer) Jn, (er) Bek, (eer) 
JFr. (ser) Sh — (©or) vg. 



hara 
hering 
sneare 
5 seer BQ (?) 
■oaru 



hare 
hering 
snare 
share 
• share 



' tonsare.* landscearu * territory * Grein, 



8t»r 

— also Btare. 

starian 

sp0r(stan) 

10 spartan 

-wBsr adj 



sterling 
Btaren 

sparen 
1*warr 



hare 


heer 


herring 


herii) 


snare 


sneer 


(plough)share 


plaujeer 


share 


Jeer 


starling 


staalii) 


stare 


steer 


spar 


spaar 


spare 


speer 


ware 


weer 



— ^war : bar prt, Balthasar Ch. 
war(e)nian wamien 



warn 



wan 



tfarenn 


flEure 


feer 


mare 


(night)mare 


naitmeer 


care 


oare 


keer 



take care/ * avoid' (aa) BU. (a) G. (00) EO, Bch, Sh, 

faran 
mare 

15 oaru 

(««>) Cp. 

tceorig tchari^ chary 

* queruloas * — ' mournful/ * sober.' tn/f. of earn. 

b»r adj bar bare 

(^eo) Cp, 

J bare 

( bore 

aUo bsr — bar Ld, iber, bear (a) Lay. bare : ])aret ber : gere CM. bar : 
iBsakar, ber : Asser, bor GE, beer : heer ( -sbSr), meiisageer, ba(a)r CA. 
bare: fare TM. 

dearr vb fdarr dare deer 

— fdftre CM. 



bar prt 



tbarr 



tjeeri 

beer 

tbeer 
bor 



FIRST WOBD-XilST. 28 I 

20 pearroo park park paak 

b»rlio Bf tbarrli^ barley baali 

Mtually b^re— bnrlio Ld, barllc Betit. barli(ch) Wid, barly CA. bere 
Ay,, Cir— (barlei) Q. («b) Ld. 

hmn baoe baas bna 

aJUk. 

arwe ar(e)we arrow »rou 

arwan 'oatapulias* Aldhgl ; frwik 01 ^r, pi ^rrw, earlUr OE lerig 

OET, earh Grein («rii) Pr. (»ro) Ld. 

■pearwa sparwe simrrow spsrou 

35 nearu narwe narrow xuerou 

gearwe jarwe srarrow J»rou 

gearwe sfpl gere gear tgier 

gfrwan, prt g^rede «6. 01 ggrvi, pi gj^rvar, vshenoe the ME g. — geren pi 
Lay. — (giir) Cp etc, 

be(a)rwe barow(e) (wheel)barrow berou 

hArfest therr&sst harvest haavist 

• autumn ' — ^a r are— e, a Td. 

30 am prt trann ran vmn 

omVPteWS. \m Ru.-—inn CM, GE, Hv, ron Kath. gam PP/, yam^y. ; 
by anal, o/eamian ete. 

geeamian emien earn een 
(«m) Cp. (mm, jam) Ld. 

ftam feme fern feen 

geam {arn yam Jaan 
(jaarn) BU, (a) G, 

earm "hamn arm aam 
(seaerm) 3ch, Sh, 

35 hearm harm harm haam 
— homi Lay., AR. 

owearm swarm swarm swom 
— — (swoonn) Beh, 8h. 

wearm "fwarrm warm wom 
(a) BU. (00) Cp, EO, Bah, 8k. 

earc sf farrke ark aak 

sroe-bisoop arch- aroh- aatj- 
— also eroe-, ene*. 

40 Btearo tstarro stark staak 

spearca sparke spark spaak 

mearo sf tmerrke mark maak 

* boundary.* meardan ' mark.' marc • mark * (coin) I WS from 01 m^rk— 

merke/remi 01 mfrki. Ch has merk ' mark * and mark * coin.' 

tb9rk bark bark baak 

mearg marou marrow msrou 
— meari Jul. mary(mery'< Ck. 



282 HISTOEY OF ENGLISH SOUNDS. 

45 eart vb tarrt art aat 

earO VP* arO Du. e«rS, eart jBa.— ert, art North, (e by in/l. ofSeand, ert). 
si^eart swart swarthy swo])! 
(swart) swart 6m, (Bwoor]>) Bwarth Beh, 8h. 

wearte werte wart wot 

— a rarer. 

teart — tart taat 

heard tharrd hard haad 
(hserU) liardly Jn. (hard) Ld, 

50 tluur]>na tharrdnezm harden haadn 

— d by anal, ^heaid. 
sweard sward(e) sward swod 

*Bkiii.' 

Bceard sherd < , ,1 faad. 

I sherd j * 

(ee) Beh, (e) Sh. 

weard sb ward ward wod 

sm * goard/ sf * guardianship.' 
-weard -ward -w^ard -wed 

(bakward) G, (bakerd) Ju, (oakird) Beh, (ookord) Sh awkward 

— (bsBkdd) etc rg, 

55 t5-weardes towardes towards t(ew)odB 

(towards) G, (touserd) toward Pr, . 

geard yerd yard Jaad 

— (ivin)ierd Ld, yerd : herd Ch — (a) G. 
beard berd beard bied 

(ee) Q), Jn. (e) Pr etc, (ee) Ld, Beh. (e) Sh. 

hearpe harpe harp haap 

soearp tsharrp sharp Jaap 

60 wearp warp warp wop 

' Btaraen/ 
^varpa warpon warp wop 

* throw.* OM weorpan * throw.* 

alu ale ale eil 

gen, alo]>. 

bryd-alu tbridale bridal braidl 

aler aUeme alder older 

air later, elren adj.— from the OS adj ; ep linden. 

65 jisal sn sale sale sell 

smaal smal small smol 

sc8b1 vb tshall shall Jeel 

Bceal WS. Boeal, seal Ba. — sal North., GE\ salle TM\ orig. a weak form. 

Bsel ^y.— (shal, au) Sm. (o) Wk. (00) Cp. (as) Mg, (Joolt) Beh, (Jselt) 
Sh shalt. 

scalu shale scale skeil 

* husk ' — also scale by ii\/l. of Fr eacale. 

scalu sooie scale(s) skeil 

* balance ' — Prompt, ; injl. of 01 skal. 



PIEST WOED-LIST. 



283 



nightingale 
tale 

dale 
aU 

hall 
BtaU 

wall 

&U 

call 



tetolwet 

wo(l)net 

wheil 

fould 

naitiggeil 

teil 

tdeil 
01 

hoi 
Btol 

wol 

fbl 

kol 



•tel-wir]>e 8talewur]>e stalwart 

' serviceable * — stalwor)> later, 

70 tval-hnot walnote walnut 

— — (woohivt) Bch, 8h, 

hwasl whal whale 

— qual : withal, hwel : wel Sv, pi fwballe North,, TM. 

&l(o)d tfiedd (8heep)fold 

n^hte-gale nihtegale 

talu tt&le 

* enumeration * (teeal) Cp. 

75 dttl td&le 

aU taU 
awl, all jETF^. 

hall Bf halle 

stall tstall 

'■iaoding.' 

wall twall 

wawl Sb, 

80 &llan tfiEdlenn 

ffiiwl HVg. fiwll Td. 

toeallian oallen 

Grein; late. OJkalla caul Sb. 

galla tgalle gall gal 

gallede galled galled gold 

'galled* {(fhor$e*), 

{tallswa also olsou 

talls(e) as »a 

— (e)a]Bwa, al8(e) Ld, alsCw)o, aae Kt. au(e) (alae) AR, ak North, fals 
'also' Ch, 

talB ttJeOls iSeOse fols 

late; from LaL or f^. 

85 salu salwe sallow sslou 

ye». salwes. 

swalwe Bwalwe swallow swolou 

(8w»lu) Pr, (bwooIoo) Bch, (swoloo) Sh, 

walwlan walwen wallow wolou 

'. (wwloo) Pr, Bch. (woloo) 8h, 

tJAltt fialwe &II0W fielou 

' pale * ; gen, falwes. 
malwe malwe mallow mselou 

90 oalu oalwe oallow kaalou 

gen. oalwes. 
half thallf half haaf 

(hoolf, hoof, hoopeni) O. (hoof) Q), Jn. (heaBpeni) halfpenny Jn. 

(hmepe» halfpennyworth Ld. (heoaef, heepini) Bch. (hmef, heepenl) 
iSA~(heipni, helpio) halfpenny, Halpin. 



al-sw& 



284 



HISTOEY OF ENGLISH SOUNDS. 



aalfiaa salven salve b»1v 

(saeav) Pr, Beh. (aoov) Cp, /», EO. (wlv) Sh—dder (aaav). 

oalf toallf oalf kaaf 

caulfe Td, (kaulf) BK. (koof) Cp, Jn. (kewf) Bcfc, ^A. 

JkalfL — oalf kaaf 

•calf of leg/ 

95 healfter halter halter holter 

ea, » 2 W8 — helfiter Som, heltir, halter Prompt, 

slmesse tallmesa alma aami 

— almus North. ; frvm 01 9liniua — (oomz) Jn. (snlmz) Bch. (flesemz) Sh. 

halm halm ha(u)lm hom 

owalm cwalm qualm fkwom 

'death; 'destruction.' 
Btaloian stalken atalk atok 

100 'walcan walken 'walk wok 

'roir (woolk, wook, walke)>) &. (wook, w»lk) W. (wook) Cp, Jn, 

Bch, 8h. 

oealc o(h)alk ohalk tjok 

— chalk Wid. chalke, calke Prompt, ch points to'Kt; cp under ctXA — 
(tjook) Q. 

bale balke 

< poroa ' WGh baloan * heaps.* 01 balk ' beam.* 

aalh Bailee 

* willow.' 
galga galwe 

(g«les) ES, Bch, Sh. 

105 talg taluh 

* dye ' — Prompt. 

halt adj thallt 

(00) Ld. 

salt taallt 
(a) Sm. (00) a, Pr, Cp, Bch, Sh. 

malt malt 
(malt) &. (00) Ld. 

aid tald 

owld HVg. (oould) G. (oold, ould) W. (ould) Pr. (oould) Jn. 

{00) Ld. 

1 10 alder-mazm tallderrmaxm alderman oldomon 

haldan thaldenn hold hould 
bowldHF^. (oou) (?. (ou) Pr,Cp. (hould, «phould) ^cA. {00) Sh. 

aalde prt tsalde sold sould 
(oou) Bll. (ou, 00) W. (ou) Cp. (ou) Bch. (00) Sh. 

fkldan folden fold fould 

oald fkald oold konld 

— chald, chold KS — (ou, oou) Sm. (oou) 0. (00, ou) W. (ou) Pr, Cp. 
(ou) Beh. (00) Sh. 

115 talde prt ttalde told tould 



baulk 


bok 


Ik ' beam.' 




sallow 


salou 


gallow(s) 


g»lous 


tallow 


taslou 


halt 


holt 

1 


salt 


solt 


malt 


molt 


old 


ould 



FIBST WOBD-LIST. 



285 



bald tbald bold bould 

(ou) 8m. (oou) &» (00, on) W, (ou) Cp. (oou) Jn. (00) Ld, Sh. 

(on) Bch. 

hra])or cp ra]>er rather raaVer 
(r»a0er) G — vg (reiCar). 

tl9)>8f — lathe leiV 

h^vrsjyer t'^he]>)>r 'whether 'wheSer 

», e Ru, — we'Ser (wajwr) Xay. wheder TM ; weder ilud. are weak formi. 

120 f)BB]»n smf fl&dme fiBtthom fieVem 

— -a28o fedme. ve)>me Xay. fkthom : com TJ£. — (ftedom) Zdf Bch, 
(fffit$dm) 8h, 

ow»]> prt towa]>)> quoth tkwoii)) 

— oweO iliZ. auo8 /u2. cod (coth) Cif. qua9, quad GJS, quod P/*/, 
CA— (ko» I'd ^kwo))) <?. (koo))) /», 5A. (l^wo))) /^cA. 

paB]> pa)> path paa]> 

baB)> sn tba)>]> bath baa)> 

(ba» bath, (baeae)?) Bath LiL (le) £cA. (seae) 5A. 

ba]>ian ba]>en bathe beiV 

125 :^ba)>a8k baaken bask baask 

' bathe oneself/ 

tma]>k ma]>ek mawk(i8h) mokij 

* maggot' — later ma(u)k. 

httsel hasel hazel heisl 

Ml twaas was woa 

— (waa) Sm, Ht, (waz) G. (o) Cp, Ld, Sh. (00) BcJi, 

tgreas- grass graas 

WS gBSTBj pi grasu. gen Da. gras- Jtu. — gras, grsM Lay, greaess pi 0, 

ffress, grin North, gera Ay. 01 gras, ODan. gr^a — (a) Q, {gnoao^r) 
Jn. (») Ld, Bch, Sh. 





grasen 


grass 


greia 


glM 


glas 


glass 


glaas 


— e AR, Ay. 












glasier 


gleljer 


{z)Ld. 








br»s 


bras 


brass 


braas 


— e ABf Ay. 








brfldsen 


tbrasenn 


braaen 


brelxn 


'of bran/* bold 


1 






135 bl»se 


blase 


blase 


bleia 


* torch; 








assa 


tasse 


ass 


aas 


(as) BU. 


(aes) Beh, Sh. 






msBsse 


tmesse 


mass 


mss 



— tmM(Be) North, fmesae Ch. +mas(Be) AllP, AuJ. — >, e) Sm. (a) BU. 

dso ash ash oj 

the tree — asche, esche Prompt. — aishe, aiss Sb, 



286 HISTORY OF ENGLISH SOUNDS. 

asoe tasskess a8h(e8) oj 

— askee Norih., AUP. assben (ainben) Ch. aiache WicL — ^OBshes Td, a 
Sm, ai Sb, 

140 Xraak rash rash reej* 

hlsBBt last last laast 

* load '^e Prompt 

wflBBcan twasshenn wash wo J 

— wetae Ay. weab 7J(f. waischen TFir'cZ. — wtd»Sb, (p) Mg;'Ld,Sh. 
(00) Bch, 

flflBsoe — flask flaask 

flaze lat€, 

xnflBBO — mash moj 

niaxwyrt late — (miij) measb EO. (as) Cp, Bch, Sh, 

145 wflBstm twasstme waist welst 

* growth ' — e Ld, Lay, 

ftost tflBusst fast flEUist 

— SB, e, a Lay, a, e North, e AUP. 

fsstan tfasstenn fSast faast 

fjBstenian ffesstnenn fEisten £Seulbii 

— e Jul, AUP, iHv, 01 festo • fasten *— (faesn) /», Sh. (fiestn) Bch. 

znflBst mast mast .maast 

* of ship.* 

m»st mast mast maast 

swina m. 

150 ^kasta casten oast kaast 

— keeten (ea) AR. casten (keasten) Jul. keste ily. kest prti fayrest 
MH. e, a CM. cast : last, kest : rest TM. f*. fe RBC. a '\Earl., 
And. e AUP. fa C'A- (a) Q. (aeae) Cp. 

oastel fkasstell oastle kaasl 

'village.* 'castle' Chr. 1052. A Winchester charter of 931 hoM iSara 
stanceastla a* a boundary — {kaeal) Ld, Sh. — (kieetl) Bch. 

best bast bast badst 

SBspe aspe aspen SBspin 

— in aspen leef (CA) aspen »* an adj; ep linden. (aspin) G, 

(flcspan) Jn. 

hflBspe hasps hasp haasp 

155 w8Bsp waspe wasp wosp 

wsefe Cp, wseps Ef. weps, weaps, wseep Wyl (a) 0. (00) Bch. («) Sh 

— wops vy. 

awel awel aw^l ol 

also &\ — eawles Kath. owel ON, pointing to OE Swel. foules Ch. el, 
pi aules AH. al(le) Wicl. 

clawn tolawmress olaw klo 

clea VP — clawe pi ON. clauen pi Ay. claw, de Prompt, clawe, de 
Wicl. — (au) Sm. 

thabban thabbenn have hsv 

bafa imper. — habben, hauen inf. Ld — (bnav) Sb. (hav) Bll. 



PIBST WOBD-LIST. 



287 



i-be-habban 




behaven 


behave 


biheiv 


* enclose.' 










160 j:h9fh Bf 




havene 


haven 


heivn 


— hteyeDe, hftfene Lay. 






hafoo 




havek 


hawk 


hok 


— haiik C\, 










hrtefii 




raven 


raven 


reivn 


IWS Xxremm, 










^alafra 




slaveren 


slaver 


slflBver 


bobSoxx 




Bhaven 


shave 


Jeiv 


165 Btof 




tataff 


staff 


Btaaf 


(a) 8m. 


(«) 


Beh, Sh, 






stafiEui pi 




tBta&88 


staves 


steivB 


waflan 




waven 


wave 


weiv 


l^vafira 




waveren 


waver 


weiver 


nafti 




nave 


nave 


nelv 


170 nafola 




nav(e)le 


navel 


neivl 


-^noule Bni. 










nafo-gftr 




naoger 


auger 


ogor 


— aUo DAvegor. 










C89f 




tchaff 


chaff 


tjaaf 


ceaf, pi c(e)afu W8 


(8b) Bch, 


Sh. 




cafer 




cheaffer 


(oook)ehafer 


tjeifer 



ceafor WS—Trevi$a. 
i^craflan oraven crave kreiv 

uncraf *d Laws of JB^, 01 krafa ' demand * ihH, kr^fja «&. 
175 {krafla oraulen crawl krol 

onalk knave knave nelv 

Scint. Generally cnapa, ' boy.* Neither onapa nor onafa in Da. and Bu. 
— knape 0. knave Lay., ^Ilv. knafe MS. p, y GE. 

g»f prt tjaff, gaff gave geiv 

ea.WS-^efAR. gaf ITtVZ. y%f Ay. yafrstafCA. ^ North , OE, TM. 
gef : )»! AllP^{g9v, gan) harhare Cp — (giv) vg. 

JgfTpfsf grave grave greiv 

OE gr»f am. 

{ thafe]>}» hath th»}» 

( ha]> has hass 

hsfeiS, hsbfiB Dtt. haef]) /?«.-> bavis (has) CM. be> Ay. bath Ch. 



h8Bf]> vb 



180 nfter 


taffterr 


after 


aafter 


rafter 


rafter 


rafter 


raafter 


-Hj^y. 








80»ft 


Bchaft 


shaft 


Jaaft 


oneft 


tcrafCt 


craft 


kraaft 


— eily. 








ged»ft 


tdaffte 


deft 


deft 


* gentle,* 


* suitable '— defte Be»t. 







288 



HISTOBY OP ENGLISH SOUNDS. 



185 hadfde prt fhaffde had h»d 

— htefde, hefde, hafde, hadde, haued Ld. hefde AR, hefde^ heuede 
(aHde) Lay, hadde:i]adde ON, hauid (had) CM\ always monoeyU. 
hedde Ay, hade : glade ^Z^P. heuede, auede, hadde (7 J^. f^adde, 
fhftde Ck. haide : saide ( nhsefde : asgde) TM, 

anoor anker 
tnjl. by Lat, anc(h)ora. 



ancleow 

— also anclowe. 
thanki 

' hasp/ ' clasp.' 
rano adj 

* proud.* 
190 hlano 
]>ancian 
sano prt 
scanca 
scranc prt 
195 Btano prt 
orano 



ancle 



hanke 



tranno 



anchor 


»i|ker 


ancle 


8di|kl 


hank 


h»i|k 


rank 


raagk 


lank 

thank 

sank 

shank 

shrank 

stank 

crank 


l8BI|k 

)>aBi|k 
s»r|k 
J»r|k 
Jr»x)k 

St£BX)k 

kr»r|k 


drank 
bank 


dr8Bi|k 
bai|k 


anger 


«X)ger 


angle vb 


aaggl 


hang 


h»r| 



lank 

t]>annkenn 
sank 
shanke 
shrank 
tstannc 
cranke 
In crancstfef — Prompt, 
dranc prt tdrannc 

{banki fbannkess p: 

01 bakki * bank of river/ • ridge.* 
tangr tanngrenn ^ 

* grief/ 

200 angel angel 

»hook.' 

hangian hangen 

ifUr, hon /r—hongien Lay., AR. hongi Ay. hing North, 

lang tlang long lox) 

o VP, Dm., Ru. o. a eWS. a lWS—\&nnge adv 0. & Ld, o Lay., 
ARf Best, Ch. o, a Ay, a North, lunj? adv And. — (logger) reciim 
(loi]»p) W. — (laeij, leii)) Lang, Laiog are North, and Sc./omu, 

ge)>rang su ]>rang throng )>roi| 

pwang t)>wang thong ]>oX) 
— thwong Wicl.f TM, thong aUo, 

205 sang 8b tsang song soi| 

— zang, o Ay. Bong : along prt Ch. song : emong TM. 

sang prt sang sang sox) 
— sang : emang TM. soong : stroong Ch. 

Strang fstrang strong stroi) 

sprang prt taprang sprang spradi) 

tvrang twrang wrong roq 
— mid wrange eb Ld — wronnge Td. 

t(a)mang among emai) 

amanges amongst emar|st 



210 on-gemang 



( 



FIRST WOBD-LIST. 



289 



inmoDg D(«. — emnftng Id, omang, (a)maiigi8, mang North, among 
ARf Best,, Harl. ameDges XJS, amang Ay, emong:a song TM 
—(o) 0. (u) Bt 

mangere mangere -monger mai|ger 

♦gang — gang g»x| 

a Scand. form whick diiptaeed the OE g^nge ' troq>.* OB gang had ofdy 
the eenee of * going, ' * gait ' etc, 

ta2kg(e) tange tongs tox|B 

}banga — bang bax| 

215 hanep hemp hemp hemp 

aUo henep. 

lane lane lane lein 

Blickl. Som. — aXeo o — {eest) Cp. 

]7anon ]>enne8 thence Vena 

— )>anony ]^o)nen Ld, ^nene, ))(e)onene, )>(e)onne, ^nne Lay. ]x>n- 
ne : monne ON, ^nne Kaih, Jwnnee Ay, thennes Ch. thine MH. ancd. 
q/*heonon. 

swan swan swan swon 

{»)Beh. {o)8h, 

wanian wanien wane wein 

(ee) Cp. 

220 hwanon whennea whenoe whena 

— hwanene, o Lay. h^eon(e)ne Jvd. bwannes Ay. mrhennea Ch. anal, 
ofhieonoa — (i) Mg, 

&na vane vane vein 

* banner * — Ch. fime Prompt. — (faan) ' weatheivock * Sm. 

manu mane mane mein 

{ee9) Cp. 

manig i'mani({) many meni 

o Du., Bit, a, e, e IWS — ^moni Lay., AB. maaj AIIP, Aud. man! Ay,, 
North, many Ch, ai$o meni ; ajtal. of Snig (?)^(a) 0. (e) C, (masne) 
tometimee Jn. (mm) Ld, {m) Boh, (e) ah, 

oran oran(e) crane krein 

— cnm Lay, orane injl. qfSeand. irani t 

gsnit 

tbein 

rsnattk 

Bpfldn 

apttn 

twon 



ten 

mmi 

kttn 

luen 



225 ganot 




gante 


gannet 


— Prompt. 








bana 




bane 


bane 


trannaaka 




ranaaken 


ranaack 


apann prt 




ipan 


span 


apannan 




spannen 


apan 


230 wann adj 




wan 


wan 


(8B) Beh, 8K 






feum 




&n 


&n 


' winnowing- 


fan.' 






Tntt.Tiii 




fmann 


man 


cann vb 




tcann 


can 


canne 




canne 


can 



u 



290 



HISTORY OP E.VGLtSH SOUNDS. 



235 be-gann prt 
taTinian 
panne 
gebann 
an-fllt(e) 



tbig^anli 

tannen 

panne 

(i)ban 

anvelt 



began 

tan 

pan 

ban 

anvil 



big»n 

tAn 

p»n 

ban 

anvil 



— anefelt, ftnefeld, anfeeld JTicL also anvylt. 

240 antefh antem anthem sn)>lm 

0, t) Ld. 

i^vanta waaten want wont 

— ^wontin Jul, — (00) JBeh, (o) 8h. 

and tannd and and 

— ant Marg,y I^ay, an Lay., OE, Ay, — an(d) Td, (een) Ld, 
and-swaru tanndsware answer aanser 

ondsuere aec. Du,, andsuari Bu, andswarian, -orian vb WS, ondaweorian 
VP. o, e Du. a, se, eo Ru. — anndswere <tUo 0. en(d)swere, en- Lay, 
auDswere Wicl, — (answer) non (aunsuer) G, (aenaer) Cp etc. 

hand than(n)d hand hand 

— o Lay., AB, GE. a, o Ay. a North. 00 Wicl. o, a Ch — a, o Td. 
(hsBnuem) ' handsome * Jn — (fasnsdm). 

245 handel thanndlennvb handle handl 

(hsenl) Jn. 

land tland land land 

— oa Prod. 0(0) fTic/.— londe Td. (ae) Cp, (laenlord) 'landlord* /»— 
(Isenlod). 

sand tsand 

standan tstanndenn 

strand tstrand 

250 iv9nd twand 

(8b) Beh. (o) Sh. 

wandrian waadrien wander wonder 

(woondir) Bch, (wonder) Sh. 



sand 


sand 


stand 


stand 


strand 


strand 


wand 


wond 



oandel 



oandel 



(luenl) • candle ' Jn. 

gandra 
band sn 



255 brand 
hamor 
lama adj 
itsami 
scamu 

260 stamrian 
firam 



( oandle kandl 

\ cannel(coal) kanl 



gandre 

tbandess pi 
Ddis Wid. 


gander 
bond 


gander 
bond 


brand 


brand 


brand 


hamer 


hammer 


hamer 


lame 


lame 


leim 


tsame 

tshame 

stameren 


same 

shame 

stammer 


seim 
Jeim 
StaBmer 


firam 


from 


from 



a Ep., IWS. o Di*., l?a.— a Ld, Kath., fHty Ay. 0, (a) Lay. o, a Ch, 

nama tn&me name neim 

(ctfe) Cp. 



FIE8T WOEIVLIST. 



291 



gftinan gamen game g«im 

— gomen Lay,, AR, geina, pi gemenes Ay. igftme Ch, gam TM. 





hAsun fif hamme 


ham 


h»m 


265 


ramm tramm 


ram 


raun 






swam 


Bwann 




(ft) 0, (8b) Bch, 8k. 








crammian orammien 


cram 


krttm. 




— cremmin Prompt. 








*atampian stampen 


stamp 


8ta»mp 




pil-BtAinpe * pestle.' 








*craxnp orampe 


oramp 


kramp 




oroiiipeh^ • folUlifl/ WGh 






270 


lamb tlamb 
Umbe Td, (lam) 0. 


lamb 


Uam 




waxnb sf twambe 


womb 


wuwm 



— wombe Lay,, AR, Ay. , AllP, Ch. wambe North, wame : came CM 
—(womb) Sm, (wuam) Bt, Cp, Bch, Sk. (aum) Jn. (woom) DyeKe, 

oamb toamb oomb koum 

(koom) 0, (kumn) Jn, JEO, (koom) Beh, Sh, 

acan aken aohe eik 

(aatj) BuU. (hedaatf) G. (ssaek) vb Fr, (eek) Bch, Sh. sbtt. «oe, ME 
eche. Mn ache, which itm formerly pron, (eitj), t« a blending of acan 

aBoer aker acre 

275 Aoem sn akem aoom 

— aconi Ch. if^. cj^corn— ^8e«kem) Cjp. 
race rake rake 

:iak lak(k)en vb lack 

* defective.' 
lacu lak(e) lake 

Wore, charier of 944 — h^from Fr lac. 
)mbo )>ak thatch 

blending of >ak and the vb ))eccben /rom OE ]>90can. 



eiker 
eikon 

reik 
l»k 

leik 



280 aacu tsake 

< strife.' 0/ 119k < sake.' 



seik 





ransaken 


ransack 


ransask 


alao 


slak 


slack 


slAk 


alaoian 


•laken 


slake 


slelk 


* grow slack.' 








snaoa 


snake 


snake 


snelk 


285 Boaoaxi 


shaken 


shake 


Jelk 

« 


floeaool 


scakel 


shackle 


•ff 


Btacu 


stake 


stake 


Bteik 


sprse prt 


tspaoo 


>pake 
spoke 


tspeik 
spouk 



spoke hy anal. ofpU sprocen. 

U % 



292 HISTORY OF ENGLISH SOUNDS. 

on-wacan waken wake weik 

290 &-w«0]iian twaocnenn waken weiken 

twlaou leuke luke(warm) luwkwom 

— wlech Jvl. lake Lay. Iheuo Ay. blending of wlacu * lukewann ' and 
g^hlSow ' Bunny/ ' warm.' 

Jflaki flake flake fleik 

(«a) Cp, 

naood fnakedd naked neikid 

macian tmakenn make meik 

— imper. mace 0. 
295 maoode prt tmakedd made meid 

— ^makede 2x2, Lay., AR, made : brflkl PC. maden pi, imaked pte K8. 
made, yinad Ay. made (aa), ymaad Ch. 

^kaka cake oake keik 

oraoian oraken orack kr»k 

'leaound* (kzmakt) 'spUt' O. 

cwaoian owaken quake kweik 

ttaoan ftAkenn take teik 

Chron. X072. io betcBoe (inBeqaor feras) in JSIf. CoU. about 1000. 01 

300 b890 tbaoo back baBk 

— o baooh «i on bee adv 0. 
baean tbakenn 

brae prt tbraoc 

blsBO blak 

gelflBoean tlaochenn 

* seize' — Litche <& Prompt. 

305 Bttcc sak sack sak 

; JTS saec— sek North., AUP, TM; 01 s^kk. -^akke Ch. 

wflBooe twecohe watoh wotj 

—a C*— waitcbe Sh. (00, o) Cp. (o) EO, Ld, Sh. (00) Beh. 

fluc sf tare axe SBks 

aga S&. (akB) Q. 

«zl axel axle aksl 

* shoulder.' »z = * axle * — infl. of 01 9XUII or Fr ai«el I — (ekstri) * axletree' 

Cp. 

wadxan twax(x)enn wax tweks 

— wAxe : axe(e) Ch. a TM. (iwoxen) ptc Lay. 

310 wax wax wax weeks 

eTd. (») Ld. 

flttx flax flax flake 

geeah prt tsahh saw so 

ea W8. ae VP. se D«. oe, ea Ru. — Beih Lay^ AR. saw North, sag 



bake 


beik 


(brake 
( broke 


tbreik 


brouk 


black 


blask 


latch Bb 


laatj 



QE. 103 AllP. Byz Aud. ze^ Ay. tsagh TM. i-Migh, fsay CL 
Btde, ei^ Wicl— (aau) Sm. (soo) &, Cp. 

dre(a)hnian — drain drein 



FIBST WOED-LIST. 293 

nhts tehhte eight eit 

— tthte Li, Lay. ahte (e») Jul, eihte (ea) AR. ^te Ay, a North,, 

Tiir— (aixt) Q, (sit) Pr, (ee, tm) Ld, 

315 eahto)>a tehhtennde eighth eit]> 

— enhto^, eihto^ AR. egtende ^y. ahtaad North, egtende OE, 
a^t^e ^2/P. ei^tith, ei^ Ifiol.— (alxt) (?. (6e)>) Beh, (eet)>) ^^A. 

hleahtor lahter laughter laafter 

(lauhtor) 8m. (lootor) /f»— (laaft/er) tfjr. 

^Blahtr alahter slaughter sloter 

OJslfttr. 

fbht prt fSftht fought tot 

^a ^or<A., AllP. &ught CA— (&uh() iSm. (foot) Jn. fought /ram ^<o 
fohten. 

m»hte prt tmihhte might malt 

m VF, Du,^ Ru, ea, n, e eW8. i IWS^a^ Jul. a (i) Lay, i AR, 
fi, u, o North, mooote : doathe ( « dohte) Hv, -fy, fa TTfaf . myghte Ch, 
— ^mioht HVff. (moot, mod) harhare Cp. 

draught I 

draft 
OJdrfttt {drooi) Cp,Jn,EO. (dmmft) Ld. (droat)^cA. (dnnit)iSA. 

tag! taghe awe o 

OE ^gb—alto 6530 0, eie, asie i>{. 61(3)0, eie (eaye) Lay. aw North, 
eige, fftge G^. foye, +awe BI^C7, awe Prompt, Ch — (an) Sm, (ao, 
aa) 8b, (oou) Q, 

l^ga en awene awn on 

* hoskfl.' 

hagu-])om hawe])om hawthorn ho]>on 

h»gl hail haU heil 

hegel VP. bagol TT^— hasel (hawel) Lay, bail CA. 

325 lagpi-t flajS lay l«i 

(lai) a— (Idd) «^. 

tla^u lawe law lo 

late OB, from Stand, pi neut. *lagu (0/ l9g) (laau) Sm. (lauful) 

BU. (loou, bofttl) 0^. 



320 :draht draht T^ \ d^aaft 



ftt-laga 
at-lah 



} 



utlawe outlaw autlo 



sage sawe saw so 

'serra.' 
sagu sawe saw so 

« »ying.* 

330 slAgen ptc tslajenn slain tslein 

ae, e Ep, m, e Piut, sb, a Or. a, e jBfe, e VP, m Du, n, a Bu,-^ 
slawoD Lay,, fHv. alagen : &gen OE. ala(i)ii North. isleia AR. 
talain, felawe Ch, f^jn, futwen RBC, 

tslfia vb tsla' slay tslel 

■IfBgen pto dee Ck, 

snngl snaii snail sneil 



294 HI8T0ET OF ENGLISH SOUNDS. 

Boaga sliawe shaw tfo 

'wood.' 

twagian waggen wag w»g 

01 vaga ' wag,' vagga ' cradle * — Ay., Prompt, — (») Ld. 

w»gn twa^^n wain twein 

W8 w6n. 

335 finger tfsiSS' &ir feer 

(fair, faair, faler, fiEtaier) O. (few) Cp. (floor, feeor, fecr) Jn 

fiBgen fidn fkin tfbin 

—(tasiLayfAR, ffAgen ^esf., &J?. ffawe CA. 

gnian i &imen fawn fbn 

JfiBtgna j 
' rejoice * — fainen Lay.., North, au PPl, Ch. 



n»gl 


tnajslenn vb 


nail 


neil 


madg 


tma55 


may 


mei 


340 maga 


mawe 


maw 


tmo 


niagon 


tntLin 


mtL\r% 


mein 


{ee) Cp. 








on-g8Bgn 


t(onn){»n 


again 


ege(i)n 


on-g»gneB 


tonnjflBness 


againat 


egenst 


oiigea(g)n eWS, 


onge(a)n IWS, 


oiig8e(g)n Du»f Ru. 


0Dge»] 



e(a), eo, se Ld, ei, ea Lay. ai North., All P. fey Hv, aye, avens 
Kt. agon : on ( — Sn), agen : ben ( «= bSon), agein : rein GE. agayn : Blayn 
Ch — agaynat (ei) UV^. ageyne, agaynst, ei, agenste Td. (again, agenst, 
against) 0. (s>gen, segeen, geinst) Jn. (segeenst) Pr» (aagaen) Ld. (»gen) 
Beh, 5A— (©gin) v^. 

gnagan gnawen gnaw no 

(nhoo) Ld. 

345 tmgl taU tail teil 

(ai) 8m. {te) Cp, 

d»g tdajj day dei 

« pi dagae— dieges, dffii(g), deig, dai, pi dagas Ld, pi da^heas, da^^ess 0. 

dei, dai pi dawes AR. day, pi da^es, daies Ay. fdai, ilke^ f^au North, 
dai, pi deies JBef^.— day, dae HVg. (dai, daai, dee) O, (dni) W, 
{djte) Cp, 

dadges-tage daiesie daisy deiai 

dagian dawen dawn don 



— later dawnen. 
dragan 



{tdraghenn draw dro 

draggen drag dr»g 

— Swed, dragga. 

350 psBgel payle pail peil 

* gillo ' gU-Prompt, 

brsgen brain brain brein 

raggig ragge rag r«g 

raggie (setoaa, aetiger) Aldhgl. 

soeaoga — shag Jmg 

< head of hair * - feax Wpl. 



FIRST WORD-LIST. 



295 



tatt 


at 


»t 


th&tenn 


hate 


heit 


lat ) 
late/ 


late 


leit 


later 


flatter 
1 later 


later 
leitor 


tlate(me)st 


latest 


leitist 


tlattst 


last 


laast 



Jbaggi iMiffge 'bag bag 

355 sagde pi*t tsejjjde said sed 

wHe IWS. gd 2>tt., £«.— «»(i)d6 Ld. aeide Jul said North. 0e(i)de 
KL sayde, seyde Ch — (ai, e) O, (e, •«) Jn. 

magden tmaggdenn maid(en) 2neid(&) 

2 W8 msden maedyn H Vg, 

at prp 

hatian 

flat adj 

lato adr 

360 lator 

latest 

(iBBfBBt) Cp. (UBsli) < lastly * /». 

}»at t]>att that Vat 

ddatB^F^, 56. 

sat prt tsatt sat sat 

aUo det— Met : feet CA— late Tri. 

satomes-dag tsaterrdag^ Saturday sartedi 

also Hetrefldsg — aeterdai (setendai) CM, 

365 tskata sf soato skate skeit 

aJUk, 

water twaterr water woter 
(waater) Bll (00) Cp, Jn, Beh, 8k. 

watol watel wattle wotl 

(00) Cp. (o) Mg, 8h. (8b) Sc*. 

hwat twhatt what whot 

(w»t) bdter (whset) Jn. (whot) Msf, LcL (wmtmi) * ■omewhat ' Jn. 

fiat 
— vet Ay. 
370 firatwan 
' adorn.' 

tflat 

orat 811 
pi cratu. 

olatriaa 

gat an 

geat pi gata W8. m, ea Du. ea Bu. — iateward Ld. jet AB. yate 
A'oWA., J2/P. yate, gate TM. gate CA— (mo) Cp. 

f begat bigat 

( got got 

ea WS—hegKi, belet, beiet Ld. bijeat Xay. bijet AB. bigat: get adv 
OE. 

Jbatna — batten tbatn 

' improve.* 



tiktt 


vat 


Tat 


firetted ptc 


fret 


fret 


iiat 


flat 


flat 


tkarrte 


oart 


kaat 



375 be-gatprt 



clatren clatter klater 

tjate, gate gate gelt 



tbigatt 



296 



HISTOEY OF ENGLISH SOUNDS. 



hSBt 


hat 


hat h»t 


hntas pi Lei. gl. 








ISBtt sf 


iBLphe 


lath laa]> 


—Prompt.— (U» BU, (bb) Bch. (ib») 8h. 




mattoo 


mattok 


mattock m»tek 


380 oatt(e) 


kat 


cat k»t 


(ie)Ci>. 








gn»tt 


gnat 


gnat n»t 


(nh«t)I,<f. 








tflBtteo- 


tatered ptc 


tatteir tflBter 


tAttocon jEfcE, 


01 t9trur * ragg.' 






pradttig 


prati 


pietty priti 


' ounning * preti HVg. (e) Bt. 


(i) Sh. 




adela 


adel(ey) 


addled addld 


filth '— « Addled < 


5gg-' 






385 adese 


adse 


adae adds 


h lad an 


laden 


lade leid 


hlsdel 


ladel 


ladle leidl 


B»d 


sad 


sad aadd 


< satiated; 








sadol 


sadel 


saddle 1 


laddl 


390 Boadu 


( sehade 

( 8ohad(e)we 


shade 
shadow 


eid . 
odou 


Bceadwian xib. alto acsed * nhade/ 






spade 


spade 


spade speid 


wadan 


waden 


wade "^ 


weld 


fiBBder 


tfiaderr 


father : 


CkatSer 


—fader Ch, TM 


— fl&fcddyr HVg. a, 


aa Ck, (a, aa) G. 1 


[sob) Wh. (00) 


Jn, (taasfSii) Beh, \fwiBiSer) 6h. 






mmdere 


mader 


madder 


madder 


395 cradol 


eradel 


cradle 


kreidl 


gaodrian 


tgaddrenn 


gather j 


gast^er 


a WS. ea Du,- 


-a id. e AR. a, (e) CA— gadre Td. 




^.fgiBdre 
to 1 


ttogeddre 


together 


tegeSer 


SB WS. 86, ea 


North. — togadere (e) Lay. e AR, TM. 


i North., Kt. 


togidre (e) Ch 


. tpgedyr, together 


il«d.— togedder, th, dth Td. 


glad 


tgladd 


glad 


gla»d 


biBdprt 


tbadd 


bade 


baod 


bade Td, 








400 bladd sn 


blade 


blade 


bleld 


tgadd 


gad 


gadfly 


gaddflai 


'goad.' 








apa 


ape 


ape 


eip 


tlaBpe- wince 


lappewinko 


lapwing 


laopwii) 



PIBST WORD-LIST. 



297 



lapiaa 

'Iftmbo.* 



lap(p)en lap 



405 



sap 



sap 



laeip 



Btbp 



Basp 

— sap Ay. 

^Bkapa tshapeim shape Jeip 

OE aceppnn. ie^ y W8 — ah from OE, shippennd $h 0. sohoppinde AM. 
aiep)> Ay. ahspen CK — •b»ppe8 46 p/, Td. isiapp Mb Sh. 



■orapian Bohrapen aorape 

— aUo Bcrapen, hy infi. ofSeand. Bkrapa» 



skreip 



1 


stapol 


stapel 


staple 


steipl 




xnapuldor 


mapel 


maple 


meipl 


1 


410 :gapa 


gapen 


gape 


geip 




tapor 


taper 


taper 


teiper 




papol-8t&n 


pobbel 


pebble 


pebl 




adppel 


tappell 


apple 


adpl 




Jhapp 


hap 


(xnis)hap 


hasp 




415 IsBppa 


lappe 


flap 
(lappet 


Ubp 
IflBpit 




hnappian 


nappen 


nap 


naop 




CMBppe 


oappe 


cap 


kadp 




(»)Cp. 










tklappa 


olappen 


clap 


klap 




OB daeppetung. 










tflBppe 


tap(p)e 


tape 


teip 




420 ta»ppe 


tappe 


tap 


tflBp 




tflBppet 


tipet 


tippet 


tipit 




— typot (e) Ck. 










taemppe 


trappe 


trap 


trep 




abbod 


abbed 


abbot 


Abet 




--aUo abbod, abbots the latter hy ii\fl. of Lat. or OFr. 






•oeabb 


Boab 

* 

ahab 


scab 


skttb 


* 




shabby 


Jadbi 



}h infi. ofSeand. : Swed. skabb. 
425 orabba orabbe orab 

tgabba gabben gab(ble) 



g»b(l) 



hire dat. thire her heer 

J IWS^-i Ld, Jul. i, e North., AUP. (hure) PFl. hir(e), here:iwere 
C*— (her, i) Q. 

oirioe tUrrke ohtiroh t/eetj 

IW8 cvToe— ohirohe Jul., Lay. cher(e)ohe Kt. kirke North., Best., 
AUr. cherohe : werche, chirohe, ohorohe, kerke:«rke (-lirk vh) Aud. 
chirehe CA— tii;i)urU Sb. ;i), (u) vtl (yy) 8m. (u) G. 



298 



HISTORY OF ENGLISH SOUNDS. 



Jhvirfla whirlen whirl whe»l 

{e)JBlL (i)G. 

430 birce birche biroh beet/ 
— €U4o 6, 1], birke. 

hirde thirde (8hep)h8rd Jeped 

i,ie eW8.ylW8. eo VP. 10 Du. 10, eo /?«.— h(i)erde CA-^hepheerd 

Td. (Jepherd) O, (Jeperd), (kduhvrd) Jn. (Jepird) Bch, (Jepard) 
Sk. 

■wilian swllen swill swil 
' wash.' aho Bwillan. 

takUBn tskill skill skU 

— flchil, skil AR, aoele, skele Ay. skile (Hkil, skill), uEuchill (sk) CM. 
skylle PC. 



mll-dSaw mildeu 

'nectar' — Prompt, meldew Wicl, 

435 ttilprp ttUl 

tilian tiUen 

dUe dil(l)e 

bile bUe 

:tm iUe 

~lxiy., ON, QE, PC. iXi^pl, adv O. 



mildew 



mil4]uw 



440 Boillixig 

stiUe 

wiUa 
wile vh — 

billsn 
• sword.' 

fylmen 

445 seoloo 



til 


til 


till 


til 


dill 


dil 


bill 


bil 


ill 


il 


BhilUng 


filig 


still 


Btil 


will 


wil 



shillizig 

tstille 

twHle 
i woll Td. (w«l) haihare Cp. (woont) won't Ld — (wount). 

bil biU bU 



fylme 
silk 



film 
sUk 



film 
silk 



aeolcen, silcen tidj. 01 silki — leolke dat. Lay., AM. aho 8elk(e). 



miloh 



miloe adj mylohe 

tgxinde-swilge gnmdeswilie groundsel 
later gnindeswylge. 

hilt(e) liilt(e) 

spildan spillen 

IdFP. \\Or.,Du. « destroy.' 

450 Wilde twilde 

milde tmilde 

oild tehUd 

(ei)(?. 

oildru pi tohilldre 



hilt 
spill 

wild 
mild 
child 

children 



mUj 
graunsl 

hilt 
spil 

waild 
maild 
tJaUd 

t/uldren 



oUd Cp, VP, Past, Or., Du. cild(ru) jBfcH. dldru i?«.— cWldren Jvl^ 
Ay., Ch. childer NoHk., AllP, Jif— (tjilren) Jnr-older (tjUdrin). 

tgildi en gUde guild gild 

•tribnte, feast, guild.' geld 'payment' OlP—^eld, jilile •tribute.' glide 
Lay., gyylde Prompt. * guild.' geldehalle (silde-) * guildhall ' Ch. 



FIRST WORD-LIST. 



299 



455 wnlf 
wl]> 



i*wiJ>J) 



smith 
with 



wyth, wylhout HVg, (w/», (yr%y) barbare Cp. 



wl}>i 



withy 



amip 
wis 

wiVl 



fiddle 
kid 
pith 
cinoe 



wi]>ig 
wl^^t) ' band.' 

fl]>8lo fl]>8le 

^)> BD tkide 

460 pi]>a P^}>'®) 

si^jMui tBi)>)>enn 

— Aeo]»]WD» u Lay. leype Ay, d)>(eii), rij^enes PP^ lithen (vyn, Mthens) 
CA. biJh;!!, sen RBC, sythen, ayn Titf — ■yns HVg, 

Bxni]>]>a sini)>]>e smithy 

— aUo vaA\\ffrom 01 uni])j» ihnmgh *8ini)rige. 

is vb tiss is 

yi, U HT>. y 5&. (i«) G^. 

his thiss his 

hya, his SVg, 

465 risen pic tiisenn risen 

)>is(s) t]>ias this 
AAy%EVg,Sb. 

— ]>i8e, ]>ese these 
(Ciiz) 0. 

ffise yls yes 

cp nese 'no.* tse, eee Du. — ^yus Lay. yet Shoreh., CM — (jis, e) Sm. 
(jiis) Af^. (is) Jn. 

wlsnian wisenen 

470 s^lysnian glis(se)nien 
(glisn)/». 



fidl 
kid 

Pil> 
sins 



smiSi 

is 

his 

risn 
Sis 



Jes 



(ed) 



glisten 



wisnd 
glisn 



tmissenn 
tblisse 



miss 
bliss 



mis 
blis 



rush 
fish 



raj 



miftffiHi 

bliss 8f 

blysB EVg, 

rise risohe 

— a TVl, Prompt, resse Ay, 
fiso fish 

— fiflsk 0,ffwn Seand. fisk. 
475 miseian mixen 

—from ^mixiftn. 

diso dish 

blsoop bisshop 

— biHSoop 0,from Bcand. biskup. 

])istel ])istel 

thysUiSft. 

wlstUan hwistlen 

later hwiiUiiui by anal, of hwlnaa, hwispri«i — (whistld) pH G, 

480 wrist sf wriste wrist rist 
(risbsendf rizbsen) wristband Jn— (riibond). 



dish 
bishop 

thistle 

whistle 



dl 

bi 



•P 



}>isl 
whisl 



300 



HISTORY or ENGLISH SOUNDS. 



mi0t 


mist 


mist 


mist 


mifltel-t&n 




mistletoe 


misltou 


tSn'twig* — onlff 


mistil. 






gristle 


gristel 


gristle 


grisl 


twist 


twisten vb 


twist 


twist 


'rope/ 








485 distflsf 


distaf 


distafif 


tdistaaf 


wlisp adj 


lispen vb 


lisp 


lisp 


hwisprian 


whispren 


whisper 


whisper 


orisp 


crisp 


erisp 


krisp 


' curly-haired.' 








lifer Bf 


livre 


liver 


liver 


490 sife 


sive 


sieve 


siv 


c^ae 8h, 








wifel 


wivll 


weevil 


wflvl 


— ciUo wevil, 6y anal, of wefan ? — (wiivil) BL 




olif sn 


cllf 


cliff 


klif 


glf 


+(5)iff 


if 


if 


IFP. i,e2)i*. 


iiZtt.— 3ef JuZ. yel 


IKt. if(yif)CA. 




drifen ptc 


fdrifenn 


driven 


drivn 


496 I}>rlft Bf 


J>rift 


thrift 


J>rift 


siftan 


siften 


sift 


sift 


swift 


tswifft 


swift 


swift 


tskifta 


tshifftenn 


shift 


fift 


aoyfte Chron, Laud 1046. 






Borift Bf 


tshrime 


shrift 


+/rift 


500 gift sf 


Jlfte 


gift 


gift 


gift (pretiam) Law of Ine ; 01 gift — ^yef|>e Ay, 




tdriftsf 


drifte 


drift 


drift 


sinoan 


tsinnkenn 


sink 


8ii)k 


slinoan 


sclynoen 


Blink 


sUi)k 


Borinoan 


shrinken 

• 


shrink 


/rii|k 


505 stinoan 


tstinnkenn 


stink 


stil|k 


winoe 


wynohe 


winch 


winj 


wincian 


winken 


wink 


wigk 


iSh, 








wrincle 


wrinkil 


wrinkle 


rii)kl 


wrinklfifft. 








fine 


finoh 


finoh 


finj 


(fin J) Rnch G, 






510 twlnolian 


twinclen 


twinkle 


twiijkl 


twiiikl56. 








drinoan 


tdrinnkenn 


dHnk 


driijk 



iSb. 



FIRST WORD-LIST. 



301 



ring 
tringenn 



hring 

ring /n'y. 

(h)rixlgan 

iSb, 

]>ing iping 

thing HVg—{nB&D) nothing vy. 



515 Bingan 

iSb, 

Bwingan 
Btingan 
spxingan 
wriBgan 

520 finger 
oringan 



tBlngenn 



ring 
ring 
thing 
fling 



Bwingen Bwing 

tstingenn sting 

tflinlngenn spring 

wringen wring 

finger finger 



rii] 
riij 
J>ir| 

Sil) 

swig 

stii) 

sprix| 

rii] 

figger 

krin^ 



cringe 

— crenchen (erangen) If ar^.— cringe it a hUnding qf the ttrong oringan and 
a weak ^cr^ngwi. 

olingan dingen kling klii| 

' wither.* 

bringan tbrinngenn bring brix| 

i VP, IWS, e Dtt., elKt. i, e Ru,, eW8^% Ay, 
•in tinn, i in in 

—alio ine Jul,, Aff ; from weak OE innan— yn HVg, Sb, 
525 linetwige — linnet Unit 

sinu sf sinewe sinew siivfu 

WS, §eonu Cp—t Lay. eo Marg. e QS. i Prompt, \, e WUt. 

spindl 

tin 

in 

skin 



soinu 




sohine 


shin 


spinel sf 




spindle 


Qpindle 


— — (sptnl) . 


In. 






tin 




tin 


tin 


530 inn adv, sb 




+inn 


in, inn 


tskinn 




tskinn 


skin 


■ojnnon Ckr. 


1075- 






spinnan 




spinnen 


spin 


g^winnan 




twinnenn 


win 


— — wynning HVg. 


wynn 8b, 




finn Bin 




finne 


fin 


535 cinn 




ohin(ne) 


ohin 


be-ginnan 




tbiginnenn 


begin 






ttwinn adj 


twin 


binn sf 




binne 


bin 


winter 






winter 


540 flint 




flint 


flint 


minte 




minte 


*mint 


hind sf 




hinde 


hind 


be-hindan 




tbihinndenn 


behind 



sx>in 
win 

fln 

t/in 

bigin 

twin 

bin 

winter 

flint 

mint 

haind 

bihaind 



302 



HISTOBY O? ENGLISH SOUNDS. 



hlndrion 


hindren 


hinder 


hinder 


545 rind sf 


rinde 


rind 


raind 


(win) Jn. 








Ilnd sf 


linde 


f linden 


tlindir 



( lime laim 

linden <idj — linde tree became linetree, Khenc€ in the tSth cent, lime- 
tree. 

Binder sindir cinder Binder 

cinder by eo7\f, vnlh Fr cendre. 

wind twind wind wind 

wynde, wyndda Td. ij Ck. (ei) (?. (ai) C/>. (winmU) JBch, 

(waindmil) Sh. 

windan twinnde(olut) wind waind 

550 fvind-fts windas windlass windles 

* winding-beam* (windlses) Ld, (wiiili») Bch, (winles) Sh. 

j;vind-9iiga windowe window windou 

'wind-eye' — eie]>arel (windohe) AJR. windoge OE. wyndowe : ycrowe 
Ch. 

windwian twinndwenn winnow winou 

findan tfindenn find fEiind 

grindan tgrindenn grind graind • 
(greinston) Jn — (grainsten) grindiitone, v^ (grinstan). 



bind 






555 bindan tbindenn 
ij Ck. (oi) a. 

blind tblind 

him thimm 

hym UVg. 

rima rime 

lim sn tlimefls pi 

(lim) 8m. 

560 Boimerian shymeren 

tymriendes (cerolei guigitis) BoQl. 

tnumol nimel 

' oapax ' — infi. ofvb niman ' take.' 
swimman swimmen 

grimm i'grimm 

dimm dim 

565 impa ympe 

'graft; 

^olimban tolimbenn 

dom prt Or. clumben pri pi Chr 11 
timber ttimmbredd p 

io +ico, i 

negtiig etc Dv, — ic, fi North, hie KS. ich Jul, AR. fii thee'oh : beech 
CA—^i, i HVg. ei Sb. (ei) non (ei) G. 

Bicol sikel sickle sikl 



blind 


blaind 


him 


him 


rim 


rim 


limb 


lim 


shimmer 


Jimer 


nimble 


nimbi 


swim 


swim 


grim 


grim 


dim 


dim 


imp 


imp 


climb 


klaim 


(klaim) 0. 




timber 


timber 


I 


ai 



FIRST W0BD-LI8T. 



303 



570 8tio6 Btioho 


atitoh 


stitj 


<8titch(in8i'de).' 




w 


stioian atilden 


stick 


stik 


'pieroe/ 'adhere'— Bteke fMt Td. 






geetrioen pic striken 


stricken 


tstriken 


wioe wiohe 


wioh(elm) 


wit/ 


wiou twuke 


week 


wyk 



also wucu i» W8 — wike Lay, wuce Ld, woke Ay, wowke, wyke (e) 
CA-(ii) Sm. 

575 flcol fikel fickle flkl 

flioorian flikeren flicker fliker 

< flutter.' 

micel txnikell much mat/ 

J IW8 hy anal, of lytel— mikel North,, Bttt,, GE, iHv, fRBC; J^om 
Seand, mikil. muche(l) Jul, much, makel AllP, inichel K8, moohe 
Ay, muche(l) (o) Ch, roekyl, moohe Aud, mekylle, myoh TJf— (a) 
8m, a. y i^6— (mitj9l) MitcheU. 



oicen ohiken 


chicken 


t/ikln 


owic tcwlcc 


quick 


kwik 


IWS cacu. cucum (and cwiouin) also in PaM, 




580 tioia tike 


tick 


tik 


pic pich 


pitch 


pit/ 


prician priken 


prick 


prik 


pricel prikil 


prickle 


prikl 


licoian likken 


Uck 


lik 


585 Ibikar biker 


beaker 


b^ker 


)>iooe )>ikke 


thick 


Irik 


— kk^oiii 8cand. ]>ykk. 






sticoa stikke 


stick 


stik 


_, / wioche 

(twikke 


witch 


WltJ 


wicked 


wlkifl 


— wioci red < bad advice ' Ld 1 140 ; 


wikke inO*^ 


trorthlew/ 'feeble;* %i^. 


o/wfiot wyckede Ay, fwikke, wikked Ch, 




flicce flicche 


fllt^ 


flit/ 


590 giccaa ^icchen 


itch 


it/ 


twiccian twicohen 


twitch 


twit/ 


— the och poxnU to *twiocan. 






bioce bieche 


bitch 


bit/ 


gesih]> sf tsihhte 


Bight 


Bait 


~i8ih-8e, siht Lay. 






be-twix bitwise 


betwixt 


tbitwikst 


i Ru, eo, u Pa$t. y IW8. 






595 wiht Bf, sn -fwihht 


f wight 
whit 


twait 
twhit 


gewlhte sn twehhte 


weight 


welt 


— wiht Lay. wy3te Ay. wyghte. 


weighte Ch, 


e due to tf^. of wegaa 



304 



HISTORY OP ENGLISH SOUNDS. 



* cany,* * weigh ' — (wwxt) G, (w«it) Pr, (wert) 3£^. (ee, mm) Ld, (ee) 
EO, Beh, 8h, 



dihtan 
pliht 
Stiffu 
600 Btigel ef 
Frige -d»g 
nigon 



dihten dight tdait 

tplihht plight plait 

stio Btye stai 

stile Btile stall 

fridai Friday fraidi 

tnijhenn nine nain 

— ^inin CM. neghen PC, ne^en Ay, neyn TM. njn Ch. 

ti(o)gole tile tile tail 

— tigel GE, t«t3ele Ay. 

twig sn twig(ge) twig twig 

twiggo, twicgo pi Dm, — twigge Pr(nnpt. tayg, tuyeggee Ay. 

605 tUegan tlin lie lai 

'jaoere.* imper, liflre— lien inf. Ld. lie North, Itgge:brigge, ftie fTv. 
tlie Ch, +lye BBC. +ly, lig Titf— (lei) O-^sf (la>- 

§ar-wiega erwigge earwig iewig 

hit +itt it ' it 
—HLd, North, hit AR, Kt, Ch'-'(h)yi HVg, liyt, hit TVf. 

]>ritig )>ritti5 thirty ])eeti 

it Du, U, t Bu,, TTi^— >riia Ld, Jvi., CM. thritty Ch. thret^ Wid., 
Prompt, thirti Prompt. 

sliten ptc sliten slit sUt 

610 smiten ptc smiten smitten smitn 

spitu spite spit spit 
* veru ' — ^i, e Prompt. 

witan twitenn (to) wit tte wit 

writen ptc fwzitenn written ritn 

wrftin Ch. (wiiitn) Bll. (writn) G. 

toitellan tikelen 

— Ch. alto kitelinge. 

615 j:glitra glit(t)eren 

glseteriim OE. 



|drit sn 
(durt) Q. 

bite 

biten ptc 

biter 



drit 

{bite 
biten 
tbitterr 



tickle 

gutter 

dirt 



tiki 

gliter 

deet 



bit 

bitten 

bitter 



t W8, t(t) VP, tt D«., Ru.^ii Jul. t Kt, 



bitol bit(t)il 

' bUtta '--aho betil. 
620 Jhitta hitten 

* find '-—Lay. 
sittan tsittenn 

vg (get). 

spittan spitten 



beetle 

hit 
Bit 
spit 



bit 

bitn 

biter 

bUU 

hit 

sit 

spit 



FIEST WOED-LIST. 



305 



gewitt iwit wit wit 

flttsf fit fit fit 
'Bong.' 

625 hider thiderr hither thitJar 

hydder Td, (h^er) Bt (tdiSer) O. {9)Mg. almost short {e) Ld. 



bed-rida bedrede bedridden 

also -reda bedreed Ck. 

riden ptc riden ridden 

hlid an Ud Ud 

]>ider t]>iderr thither 

— thider Ch — ^thyder, thether Td. almost »hort (e) Ld, 

630 Bliden ptc Bliden slid 

wldwe "hwiddwe widow 

i, u W8 — widewe AB. wydwe Ch. wodewe Ay, 

hwider hwider whither 



owidu 



r quide guid 

i tctlde oud 

hwltquidu Ep., -cadu Cp— code (quide) Wiel, 'end.' 

gidlg gidi giddy 

'InBone.' i/oryt 

635 biden ptc — bid(den) 

— bedenn ' oommanded * 0, beden (bode) Wiol, 

]»ridda t))rid(d)e third 

tJirda 2)ti.-->rid CM. thred PC. >ryde AllP, ' 
thyrd TM. thred ^«k2.— thrid, thyxd Td. 

middel middel middle 

t5-midde« amiddes (a)midst 

onmiddan — amiddeD, aiQidde(B) — in the inydd(6)B of Td. 

tbid(d)enn bid 



Irwid 
kad 

gidi 

bid(n) 

]>eed 

thridde Ch. 



tthryd 



midl 
(e)midat 

bid 



bird beed 

bred Aud. byrd:betyde prt TM-- 



biddan 

'pray.* 
640 bridd tbridd 

birdas !>».— fbrid North., Ch. 
hrydd Td. (bird, bard) G. 

•lipor Bliper 

Boip an ship 

gripe tST^V^ 

Uppa lippe 

645 Jklippa tclippenn 

ribb sn rib(be) 

tUbban tUbbenn 

he leoia)> WS, VP, Ru. lyfa> IW8. Ufe|> Du. pi Ufti)ga> FP, JR, 
Dm., Bu.'^Un^n inf. Ld. he Ufe]>> 0. Uum Aud.—lmik Td. 

god-sibb tBibb adj gossip goftip 

(p) Ch. 



slippery 


sliperi 


ship 


Jip 


grip 


grip 


lip 


Up 


clip 


kUp 


rib 


rib 


Uve 


Uv 



3o6 



HISTOEY OF ENGLISH SOUNDS. 



6 (eo). 

tse pe the tSi, t$e 

ae, ^ Du,, Bu, ^ by anal. ofym» etc dde HTg, Sb, (e) $hort Lfl. 



hert 



hart 



haat 



650 heorot 

hart 56. 

Bw^zian tswerenn swear sweer 

(fonweer) Cp, (foneer) Jn. (seer, sweer) Li. (sweer) Bch, 8h. 

smeru smere smear 

Bmir(w)«n vb — smerenn vh 0. — (ii) Cp, Mg. 
Boeran sheren shear 

sderan, y W8 — — {ee) Cp, Cjiirs) shears Cp, 
spere spere spear 

(ee) G. 

655 wfr 8m were wier 

twer-mod wermod wormwood 

wf rian weren wear 

' wear dothes ' (ee) Cp, 

fi^rian ferien ferry 

' carry * — feri * ponto * sb Prompt, f(pja ^f * ferry * 01. 

mf re em mere mere 

660 m^re ef mere mare 

— ^mere, mare Ch. 

te(o)ra tere tar 

scipt(e)aro Leeehd.—ier OS, tar 7Jf — (ier) Cp, 
teran teren tear 

{ee) Cp, 

teorian tiren tire 

eo, y IW8, eo i?i«.— eo, i. 
peru pere pear 

665 bera bere bear 

(baar) BU. (seee) Mg. 

beran fberenn bear 

(ee) a. (ee) Cp. (ee) EO, Bch, 

steorra sterre star 

— steorme from EScand. sljema. steme North., AUP, fflr, •fBBC. 
stamys : thamys TM — starre Td, sterr Ck. (star) 0. 

feorr t£B(o)rr flar jkar 

m^rran marren mar maar 

— ^me(a)rren Jul. merre, mired North, a AUP, And., TM — (a) Sm. 

( ohar ) tjaar 

( f^ar J ed^aar 

* tarn,* ' time *— chearre, chere Jul, cherre AB. chaire inf: waire ( a wsef 
adj) MH. f cbaren OE. cayre, fdiarde prt AUP. 01 keyra. 

eorl teorl earl eel 

— eorl, earl Ld; infl, ofScand, iarl (?)— (eerl, erl) Q. (serl) Cp. (ee) Ld 



smier 

Jler 

spier 

wier 

weemwud 

weer 

feri 

tmier 
meer 

taar 

teer 

taier 

peer 
beer 

beer 

staar 



670 c^rr 



char 



FIRST WOED-LIST. 307 

ceorl foherl ohurl tfael 

— oheorl Xay., AR. foherl Ch. e, (u) PPl, carl, ohorl AllP. i Wiel, 
01 karl— (u) Bll 

oftHo - carloo oharlook t/aalek 

eoTpe te(o)r)>e earth ee]> 

— ur)« ilttP. er|>licbe, yorJ» i4y.— yerth ^F^. Cer|>) &. (oer]>) JBW. 
(^»} (j^» ^ctrhare Cp, (jer]>) pa« <^ bd U9ag9 Mg. (e) X<{. 

675 heor]> her]i hearth haa]> 

(e) a, (») Cjp. 

weor)> twurr}) worth wee]> 

w6or)> «6. u, y ZIF^. o Du. eo i7«. 04/ weor^, wier^ W8, wyr^ 
ZTTi^. wor)>, wyr)w i}tt. wyr^ i2u. — wnr^ a^' i(i{. wi]r|> o^;' A a<Ji. 
wor]> #i, a<<; X< — (u) Bll^ O, (pen«r)») 'pennyworth* Jn. (uu, ») 
iTO, £cAy iSA^(pene]>). 

weor]>-Boipe twurr]>Bhipe worship weejip 

— wur(0)0cipe Ld. worsdp (i) CM. wor(»B8ipe Ay. — (wurjlp) 0, EO, 
(«) Bch, Sh. 

tgJ9r)> ger]> girth gee]> 

}>er8can tjmsahenn |^jy^^^| >'»/ 

(e) jBW. (u) harbare Cp. 

680 )>er80old ]>reshwold threshold ]>re/ould,-eld 

PaH. ^xwold, )>nd-, P>^oo-» ^rzwold 2afor. 
fidrso fresh fresh frej 

m^rsc mersh marsh maaj 

(ninj) /«, Ld. 

berstan tbresstenn burst beeet 

— beontan Ld. bresien f North., AUP, fCh ; from Seand. breeta. bersten 
AB. thryat, tbrert, fbrart TM. 

sweor&n swerven swerve sweev 

' file, nib off ' (»werv, a) Q. 

685 steor&n sterven starve staav 

* die of peetilenoe.* 

oeor&n kerven oarve kaav 

oer&Ue ohervelle chervil tjeevil 

eoman te(o)menn run ran 

rinnan 'coagulate.' eoman VP, Du. ie, i eW8. y IW8. rinna, « 07 
^-eornen AB. eomen, imen» omen Lay. yemen Ay. rin(ne) North. 
tryn TM. renne(n) Ld, iBest., iHv, MIP, Aud., fCh. ninnande 
MB — runne Td. 

eomest emest earnest eenist 

(ee) G. (ee) Cp. (ae) Ld. 

690 leomian flemenn leam leen 

(ee, e) O. (ee) Cp. (e) Ld. (Iseemii)) Seh. (IvmJi)) Fr. (lemii)) 8h. 

(vnlaeaemid) Beh, (vnlsmd) Fr, (vnlemid) Sh unlearned— (laan) vg. 

st^me adj tstime stem steen 

ie, y IFiS— stume BOl. Bi(i)eme Ch. 

X 2 



308 HISTORY OF ENGLISH SOUNDS. 

geom adj t5e(o)ni6im yearn Jeon 

girnan 06— geomen, isroen Ld. yemen, y«men Tif— (jiirn) EO. (e) 
Bcht Sh, 

bf m sn tberme barn baan 

Bb^r»-sm ' barley-houBe.' ber^m Du. — heme : yeme CAr— (baam) BIL 

beoman intr ) f tb»menn ) ^ 

. / K > bum been 

badman tr j ( tbrennenn J 

ifUr bdornan VP, Du., Ru, i(e) eW8; y IWS — eo Lay,, Marg. birne 
Mom. brinnen North., W^* bame TM. The tram, and intr form* 
are confused in ME, Injl. of 01 bruma inlr and br^nna irane. — bume, 
y, burnt, brent Td. (u) Bll. 

trans, e VP — ^beman, » Ld. be(a)men Jul. brennen tHv, AUP, CK 
bronne, brinne North, 

695 beorma • tberrme barm baam 

sine(a)roian smirken smirk smeek 

weore twerrk work week 

w(e)oPC IW8, were VP, Du. we(o)rc, w»ro Bu. — we(o)Pc Ld. were 
(o) Lay. were Jul. werk fNorth., AUP, fHv, +CA, E8, work Ay. 
wark TM— 00 Ck. (u) Bll. (aa, «) EO. («) Bdi, 8h. 

deoro derk dark daak 

—a, (o) Jul. a Marg. (dorck) Lay. (u) PPl. e AUP. fe, i Ch. 
a Aud., TM. also eo— a, e Td. e Ck. (a) G. 

beoroan berken bark baak 

700 beorht fbrihht bright brait 

-breht eWS. -bryht IW8. berht VP, Du., Bu. brehtum Du., Bit.— 
briht Jul. bricht K8. bright North., CA^bricht £Ty. 

hfrgian herjien harry thadri 

— ^hiergieDy her(i)5en Xay. herhien Kath. heri, bared North, haryen, 
harewen Ch. 

t^rgan tarien tarry tflBri 

'torment' — ^y IWS — ^terwin Prompt, terren 'provoke' Wiel. targi K8. 
ftarien Ch. 

dweorg dwergh dwarf dwof 

— also dwerwe, dwerf— (dwooif) Bch, Sh—vg (dwoft). 

beorg bergh barrow b»rou 

' mountain ' — berb3e (borewe) dot. Lay. 

705 bfr(i)ge berje berry beri 

heorte therrte heart haat 

— therte Ch. hert Aud. hart : quart T3f— herte Td. hart 8b. (e) Cp^ 
Jn, EO. (flB») Bch, Sh. 

^bvert av i*]>werrt thwart ])wot 

— ouerthwart (-twert, -twart) Ch — (over^wart) BuU. (J^vrt) Jn. 

smeortan smerten smart smaat 

heord sf heerde herd heed 

heerd Td. 

710 sweord fswerd sword sod 

eo, u IW8. o Da. eo Bu. — eo, e, (e) Lay. eo AB. fo North., Ay. e fffv, 
fBBC, tC*— 8weard(e) Td. (»wuurd, u) Bt. (>wwd) Pr. (suurd) Cp. 
(Boord) Ld. 



FIRST WOED-LIST. 309 

gfrd sf t^errde yard Jaad 

*rod* yarderef. 

■telan stelen steal styi 

(ee) W. 

"wel twel(l) well wel 

— wele:fele v6 North, wel: del (-dSl) BBC. we(e)l:deel, wheel, fel 
prtCh, 

wela wele weal twijl 

(ee) Cp. 

wf liso walflh Welsh welj 

wflhiae Kt eh as prop, name, wylieo IW8 — FPL wnlao, e (wab) Lay, 
(wolJ)Walah. 

715 weoloo whelk whelk whelk 

tfelo-fbr fbldefiftre fleldihre ^Jldfeer 

-—Ch, fe](de)!are Prompt.— (feldf«cr) Cp, (fiilfienr) Jn, 

ofle tohtte ohiU t/U 

ie, J f^£^-ohylled prt AUP, 

t]iJ9l kele keel k^l 

OEo^B — aleou, 
^lles elles else els 

720 eUe(r)]i sn (f ) eller(ne) elder(tree) elder 

^ldyr» hyldyr, hilleme tree Prompt, 

h^ sf thelle heU hel 

sfllaa tsellenn seU sel 

e VP. e, «a Ptk e eWS. y {FSf— «o, n Lay, a Jul., AB. e Kt, 
i, e Wi4il. 

BweUan swellen swell swel 

^smellan smellen smell smel 

hondsnuellM 'alapM* D%. ■myllendiiin (orepantibas) BoOl—e, u, i. 
725 sofU af shelle shell Jel 

spell Bn spel spell spel 

'•tory.' 

w^e welle well wel 

Ml fel Ml fel 

'•kin.' 

ffUaa feUen fell fel 



730 owfUaa 



. -, f quell kwel 

^""^•^ {km ui 



' kill *--H3wellen Lay., JvL oallan ' strike * Lay., AB. quelle North. onUs 
<kiU' PPL tqneUe, fkiUe 6a<4 'kill* Ch, TIT— (nuuikweler) nun- 
queller ' homioid* ' 8m, 

gellan 5ellen yell Jel 

t^Uan ttellenn tell tel 

belle tbeUe beU bel 

tdv^a tdwellenn dwell dwel 

735 melu txntte meal mljl 

* fiuina/ if en, melwes 6b impliee (meel). 



3IO HISTOEY OF ENGLISH SOUNDS. 

geolu jelw(e) yellow jelou 
(j»lo) Jn — (j«lor) vg, 

olf elf < 

I oaf touf 

m, e, J IWS; also -clfen— felf Chr-older gpeUing of oaf w aulf; 01 ftlf. 
(oo^ oof) auf, awf •/». 

Belf taellf self self 

y IW8, eo VP, Du. e(o), y J?«.— eo Jtd. u iiB. e(o), n, (i) Lay. 
e iT/, tCA— Belve, silfe Td. 

eeolfor tenllfer silver silver 

eo VP. eo, 10, nlofr, gylofr eW8. salfer Dii. sylfur Ru. — silver, eylver 
Ld. eo AR, eo, a Lay, i North,, AllP, GE, Ch, Scand. silfr. 

740 so^lf shelfe shelf /elf 

' pinnacle ' — Prompt, shelves pi Ch, 

twflf ttwellil(e) twelve twelv 

twelfe eubxtantival — ^tweolf, twelve, twself Ld, twelf, eo, ea, 89, a Lay, 
tweolf, tweolve AB. tuelf Ay, — (twdmon))) ' twelvemonfch ' Bch, Sh. 



delfan 


tdellfezm 


delve 


tdelv 


twflfta 


twelfte 


twelfth 


twelf}> 


fin sf 


elne 


ell 


el 


— alto elle. 






• 


745 elm 


elm 


elm 


elm 


helm 


helm 


helm(et) 


helmit 


helmet from 


Dutch (?) 






helma 


helme 


helm 


helm 


'clavuB.' 








swelo 


tswillo 


such 


Mtr 



BwilcsB Ep., swelce Cp. swelce VP. e eJTS. i, y IWS. e, » Du, 8b, I 
Ru. — Bui(l)c Ld, Bwilk f North,, fMv, swich Kt, swulc (solch), snch 
(o) Lay. swuch AR. swich (such) Ch, sech, soch Aud. swilk, sioh, 
such, -filyke \/rom Seand. slik] TJlf— syts HVg, (a) (?.— (dtj) vg, 

hwelo twhillo which whitj 

e Ep,, VP, eW8, i, y IWS. e, le Du. e, le, i, y Ru. — quilc North, 
hwulo(h) Lay. hwuc(h) AR, which Kt^ CA— -{hwid^, hwitj) Ldt 

750 meol(o)o tmillo milk mjlk 

mile VP, Du., Rit. — e Ay. i Xay., t-Ht?, fCA — o^^r (milk). 

geoloa jolke yolk jouk 

-HxUo gelke— (jelk) My. (jook) Cp., <SA. («) Lcf. (jdk) -BcA. 

belcettan belken belch belt/ 

ea J?/fc^, i Ru., y IT^^— Fici.. Titf. 
seolh sele seal sUl 

pi seolas. 

tswelgan tswoUjhenn swallow swolou 

prt Bwalh, ptc swolgen — ^swelghen North, zuel^en Ay. swelwin Prompt, 
swelwed (soalhid) RBC. swal^en Lay. swoluwen AR, swolwen Ch — 
(swaoloo) Beh, (swoloo) Sh, 

755 twel(i)g wilwe willow wilou 

wiligpe * basket ' — aUo weloghe. 



FIRST WORD-LIST. 



31' 



( felly fiBli 

( felloe felon 

{bellows belouB 

beUy beU 

m Ep. bel(i)g, bjl(i)g IWS'-hAy 'belly' Prompt,, Ch. often bMi. 
beli(e8) ' bellows* AR, bely Ch. beln (bdw, bely) Wicl alto beloww. 



tfiS 



belg 



felwe 



beU 



belgan 

' be angry.* 

smelt 

760 felt 

spelt 

meltaa 

b^t 
01 b^lti. 

^Idra cp 

765 ^Idestspl 

seldon 



belwen 



bellow 



smelt 


smelt 


felt 


felt 


— 


spelt 


melten 


melt 


belt(e) 


belt 


telldre 


elder 


eldest 


eldest 


tseldenn 


seldom 



belou 

smelt 

felt 

spelt 

melt 

belt 

elder 
eldist 
seldom 



—wide Lay,, AR, Ch. seldam GE. seldom PC, Prompt.— (siild urn) BU. 

scold sheeld shield J^ld 

ie, y W8, seildan vh ; ie, y W8-^ AR, Ay. e(e) Ch. e(i) North, also 
i; from viT vb Bhildenn 0. i Lay , Ay, QE. also n. e Tif— (ii) Q. 



g^ew^dan tweldenn wield 

(ii) 0. (oi) Jn. 

&ld t&ld field 

— ee Ch. ey 3W— (ii) G, Cp, 

770 geldan tgeldenn yield 



twflld 

4Jld 

J«ld 



le^ y W8-ipM, gield (yald, yefld) North.-, 01 gjalda. ^fld i^iol.— eild 
{for ield t) HVg, (j"ld) ^' (i»Wi) Ht, Jn. 



helpan thellpenn 


help 


help 


hwflp twhe(o)llp 


whelp 


whelp 


gelpan tgellpenn 


yelp 


Jelp 


'bOMt.' 






le]>er le]>er 


leather 


leSer 


(e)G. 






775 sw§]>ian swa]>in 


swathe 


tsweiS 


— Prompt, also e. 






swf )>el swe]>leii 


swaddle 


swodl 


— a^ swa^ild pte, swedyile : medyU 


ie(-middel) TM, 




;st9]>l sti]>e 


Stithy 


tstltSi 


— 8te]7i PC. rtyth : smyth Ch, 






we]>er we)>er 


wether 


weVer 


' iheep ' — ^wedir Prompt. 






fe]>er sf fe)>er(e) 


feather 


fetSer 


780 iieo)x>r ne]>er 


nether 


tneVer 


— ^ (nee1$er) /». 







312 



HISTORY OF ENGLISH SOUNDS, 



beneo]>aii tbinejienn beneath binti]> 
(biineeC) Bll. (bine))) G. (bmee«) Pr. 

be-owe]>an beowe)>en bequeath bikw4j]> 

we8(u)le wesele weasel w^sl 

besma besme besom tbysem 
(biisam) Mff. 

785 pl(o)Be peae pease p^s 

pi pioaan. 

oresse cresse oress kres 

cene, o»ne IWS — kene, fkera Ch (krijsiz) vff. 

rgst sf tresste rest rest 
reest Td. 

sweoster tsussterr sister sister 

Bwostor Or. iwustor, swyster IW8, 8(w)oe8ter Du. Bwuster, n jBic. — 
s(w)a8ter Ld. suster (o) Lay, Boster AR. soeter Ay. BUfter, fo Ch. 
Bister North., GE, TM, Prompt— u, y Td. 

west twesst west west 
weest Td. 

790 nest nest nest nest 

cest sf oheste ohest tjest 

le, y Tr5— kiate North . iHv. kyiite AllP ; 01 kista. -ftsheBte Ch. 

gest jeest yeast JUst 
(ii) 8m, (jiist, iiBt) Jn. (jest) Beh, 

g^t tgesst(hus) guest gest 

gesthuB VP, jBfeJS. ie, 7 W8. e in JBfe. from 8cand. g^st— gest Ld. 
gist (ea) AB. the gfrom Scand. — gest Td. 

geostran-cUdg gisterdai yesterday jestedl 

ie, y TTiS—^surstendai (o), gerstendai Zay.-<-(i8terdee) Jn. 

795 eowu sf ewe ewe Juw 

(jeu) Ht. (eeu) O. (yy) Bll. (eu) Bt, Pr. (juu) Ld. 

strfwian tstrawenn strew struw 

e eW8. eo J^cH—t^itewen North, strowin Prompt. Btrawen, Btrewan 
C&— Btiawe Td, (eu) Sm. (00) G. 

e&s sf sTese eaves 

— eouese Xoy. ouese Best. 

e&n tefenn even 

(iivn)6?. (liven) i>r. 

on efen onefent anent 

— onont Jul. aUo anent(i8). 
800 9fete evete newt 

— later ewte— a newt /rom an ewt. 
thfbban thefenn heave 

imper. h^fe. 

hffig thefij heavy 

(ee) G. 

heofbn the(o)fEhe heaven 



«vn 

tenent 

sjuwt 

hUv 

hevi 

hevn 



heofone/«». IWS; by anal, of eor>e— hefii HVsf. (e) Bll. (ee) G. 



1 

1 

1 


FIEST 


WOBD-LIST. 




aeofon 


tfl«fenn 


seren 


Bevn 


MtkvenCk. 


(■evn) G, 






805 wefim 


weven 


meave 


wtjv 


&&r 


fbvre 


fbver 


QjTor 



313 



— (rfw fivre. from Fr fievre (?). 

olaofian eleven cleave klJUv 

'adhere/ eo, 1 W8. eo Bu.—eo Lay. e North,, WuH., Ay, dUo i. 

geHui t{liS9nn(g) give giv 

gibaen pte Ep., e Cp. ge(o)fan VP, Du., Eu, ie, y ITi^— i^iaen, iiuen Lff. 
tgif North, geouen /«2. jeuen AB, seuon (i) Zay. yeuen Kt. 
geuen |9te: diyuen u^Z/P. fgiuen GE. syuen TFioZ. iTiuen, jeaen 
Cfh — geve* foryeven pte Td. (giv, giiv, gii) O. forgijv Ck. 

weft(a) weft weft weft 

810 hl^noe linke link lii|k 

— k/rom Scand. hl«nk (0/ hlekk). 

et^e stench stench stenj 

wr^oan wrenchen wrench renj 

flrf nciso firensh French teenj 

cw^ncan fcwennkenn quench kwenj 

— quenohen AB, Best, kaenohe, he kaenc|> Ay, prt cwelnte AB, owenchte 
(owenote) JtU. pte qneynt Ch, 

8ig dr^can drenchen drench drenj 

— dreynt pte Ch, 

bfno tbennk bench benj 

— aUo o bennche 0, bynke TM. 

l^eten tlenntenn Iient lent 

— lengten, lentedtid Ld, leinten (lenten) AB, flente Ch. 

ipngla-land engelond Bngland igg^end 

— Ch. englftland, englelimd Ld. inglandCif. yngland T3f— (ingland) 0^. 
(iiilgbend) Fr. (u, i) Jn. 

fngliso tenngllaah English ii)gUj 

— ^finglifl North. 'Hnglie, inglyeoh BBC. engleis GE. engUsBe, englis 
Ay.'~-{u^]Uh) BU, (1)0, {ii) Pr,Jn. 

820 tl^ngo leng]x> length lei]]> 

slogan eengen singe sing 

— Mind pte Ch, 

str^g strong string striij 

—^, I Ch, i FrompL — gfrom Seand. string. 

^tv^og winge wing wig 

01 ritng — ^hwingen pi Lay., AB. wingen Ay. — ^wing SVy. 
tm^ngan mengen mingle niix|gl 

825 ttrfngpo tstrennc]>e strength strex)]> 

— atrenflVe (g) AB, aUo streintiie. atrenthe North, atrengthe Ch, 

tsldngva alingen sling sliij 

geong tgnng young Jai) 

gong, inog VF. giung Du,, Bit, ging Dm. ging, iung jB«. — innge Ld. 
song, 2ing» ^eng Lay, ^ung AB. yong Ay, ging CM, yhung PC, 



314 HISTORY OF ENGLISH SOUNDS. 

yimge:tuiig«jffv. tying, o, on TAf. f^yng. gonge : tonge i2£C. +5ynge, 

o, u Harl. $ong AllP. gong : tung, e Aud. o, (e) CA— ywng JELVg, 

heonon(6) heiine(8) hence hezuB 

heonon^ IWS — ^heonne [^omheonone] (hinnes) Lay. fhenneB RBC hence : 
pence TM, hennuB }Fu;2. fhenne Ch — (bins) Mg, 

hleonian lenen lean I^n 

830 wfnlan wenen wean w^n 

(ee) 8m, (ee) Cp, 

owene owene quean tkw^ljn 

(ee)fi^ffi. (e)Ld (i) J3cA. (ee) 5A. 

geon gon yon Ijon 

geonian jenien yawn Jon 

geoniftn, gftnian — ^jeonien AB, ganynge, ganynge Prompt, alto gonen. 

pining pen! penny peni 

pending in Ki ck — penig, penegas Ld, pi pans Ay, pens, pans Wid, 
pens (pans, penys) Ch, 

835 hfnn sf hen(ne) hen hen 

tr^nnan renlis rennet renit 
' make to run, coagulate ' — ' ooagulom ' Prompt, 

wf nn sf wen(ne) wen wen 

ff nn fen fisn ftn 

mf nn pi tmenn men men 

840 Jk^nna kennen ken sb tken 
' know ' — Lay, kennenn ' beget ' 0. 

gr^nnian grennen grin grin 

df nn den den den 

* Bvine-pasture/ 

Pfnnan pennen pen pen 

C^nt kent Kent kent 

845 twentig ttwenntig twenty twenti 

^nde tende end end 
— e, 8B (ea) Lay, ee Wiel, — (iind) harhare Cp, 

end-lnfon en(d)leven eleven ilevn 

endleimn Or, SBllef- Patt, lellefno 2>u.— enleven (eoUere), SBlleven Lay, 
elevene Ch — (elevn) G, (eleven, ilsBTon) Jn, 

r^dan renden rend rend 

also hr- in Du, 

B^ndan tsenndenn send aend 

850 8^d ptc tsennd sent sent 
—sent fCM, PPl. 

spfndan spenden spend spend 

wf ndan twen(n)denn wend wend 

be-geondan t5onnd(hallf) beyond bQond 

— ^2i(e)ond; g^nd Lay, gionder (yonder) CJIf, bygonde : londe AllP. 
gund(e) QS. bigende Wicl, — (jwoder) Jn, 

b^ndan benden bend bend 
'bind,* 'bend.* 



FIRST WOED-LIST. 315 

855 bl^dan tblendenn blend blend 

' blind ' — forrblendenn ' blind.* oof|f. w. bUndan ' mix.* 
eom tcunm am mm 

6H111 VP, eom, am(in) Do, earn, (n)nm, Bu, — (e)9m Ld. am /u/., Lay., 
North., ^GE, iBBC, Kt, Ch. 

h^mxn hem bem hem 

Jst^ma ■temmen(f) stem stem 

' rtop.' • 

stfmn stem stem stem 

860 l9mp(healt) — limp limp 

recenian rekenen reckon reken 

reoon ' remuneraUo ' — ^i AB, rekeni Ay. rek(e)ne Ch* 

tleka , leken leak l^k 

OEhlecadj 'leaky: 

spreoeai tspekenn speak spljk 

specan IWS, IK, gpr ^fcU (ee) G. 

weooe weke wiok wik 

eMn weeke. 

865 t(T)r9k sn wrek wreck rek 

'anything driven on Bhore.' wnec 'actuariua' Ep., elsewhere 'exfle.' — 
Prompt, wrak (werk) Ch, 

wrecan wreken wreak trek 

tcw^oe-sand — quicksand kwiksflond 

' quake-iand *— cfp quagmire, ME quikemire. 

geoel (ia)ikel (ic)iole aisikl 

breoan tbrekenn break breik 

(ee) G, Pr. (ii) Beh, 8h. (brekwieat) in $ome eottniies Jn ; (brek- 

fist) Bch, (brekftest) /^^—(brekfeBt). 

870 rfccan freokenn reok(les8) rek 

r^oean, prt roHte, tock the place of ♦rtScan in OE; reocileas in Q>— reecho 
ON, reohe Beet, freke North, leochen, rekken Ch, 

strfocan streochen stretch stretj 

speoca spekke speck spek 

wr^oca twrecohe wi«toh retj 

' exile *-Hi4/ ytBt wreooe stede Ld. wriochid, wrechid MK, wrichede KS, 
wriche TJtf-~(retJ) Ld. 

ffocan tfoochenn fbtch fbtj 

prt fetode— ffette pH Ch. fetche, foche, ffott i^f TM. 

875 hnfcca nekke neck nek 

— nhicka Ay, 

gemfcca sf tmacche match mat/ 

< wife '— ' wife.* meche : reohe * care' BeH, meoche BBC, maoche Ch, 

bl^)>a — blight blait 

• vitUigo ' Sp, 

feoh t&hh, fb foe 4] 

— feoh (feo), faei Lay. feih AB, 

hlfhhan tlahhjhenn laugh laaf 

hlafhaC VP, hli(e)hhan Pa*<.— lehgen, lihgon (lahje) Lay. lauhwen 



3l6 HISTORY OF ENGLISH SOUNDS. 

AR, Ihejje Ay.—QAuh, laf) Sm, (loox, laf) G, (l«f, bo) Jn, (l»f) 
£0, Sh, (laBsef) Ld, Bch, 

880 Bex t8ex(tene) six silcs 

e i>tt., 3?u. ie, i P<ut,—e, eo X^. e North., Beft, fRBC, i Lay., Ju/., 

sexta teezte sixth uikBp 

— sexte North., AUP, GE, also sesia. nzte CA— eixte W. (rixt) 6?, 
Bch^ {nkM^)6h. 

reht trihht right rait 

eVP. jeWS. i iTTfif— richt BTg, Sh. 

*:}>eht ]>lht tight tait 

OJ^Stt — a2«o tight, by injl. o/'tSgan. 

*^8leht Bliht Blight slait 

01 BlStt * smooth ' {ai)Ld. 

885 feohtan ffihhtenn fight &it 

e VPf North. — eo, i Ld. ei, i, e Lay. e Jul., AUP, Aud. i AB, North,, 
GE, BBC, fCA— (feit) Sm, (feixt) G. 

n^ht tnihht night nait 

e, 8B J^. SB VP, Du. ie, i P<ut. i IWS—oUo nahht 0. niht L(i, Lay, 
nahtP*. tnightCif. ny^t, fna^t ^WP. 

mf ht sf tmihht might mait 

86, e eNoHh. ee VP, Du., Ru. ie, i Past i IWS—nUo a 0, miste CJIf. 
mught, o PC nuehte (i) Lay. mihte Ji2. migte Jy. magt ul22P. 
mi<^ : ricth Hv, tmigt GJ?. 

oneht tonihht knight nait 

e Cp, FP. eo, io, i Past, e, se Du., J?«. — cniht Ld, Lay, AR. fknight 
North.— kmiM HVg. (nbait) Ld, 

Jel tajs ay(e) tol 

' ever.* OE ft(wa) — aUo a 0. a, ib Lay. a A£. ai, aa : sua, ever and a 
CJIf. ai Be$t. 00, ai OE, ay CA— (ei) Sm. (eei, ai, aai) G, (ni) Cjp. 

890 th^ge hegge hedge hed{ 

cJm> hfcg (?), h^ (ed^) •^». 

tlfcgan tleggenn lay lei 

imper. l^ge — ^he le^^e]))), imper. le^ 0, legsen AR. lai:ai CJf. fleye 
BJBC legge:abegge (■"bycgan) C^^(lai7lee) C 

t]>eir t]>e5S they Vei 

OE hie, >5r-a^o >a 0. thai North, ])ay JZZP. thai Aud, they PP/, 
CA— ddey, ddei EYff, (ei, eei, ai, aai, e) G. (sei) Pr. (ee) Ld. 

tsfcgan tseggenn say sei 

imper. B^ge — seggon, ssegen, 88ein,'8ei inf Ld. imper, 8055 0, siggen (e) 
AR. seggen, Biggen, saggen Lay, zigge Ay. aai North, seye \RBC, 
Ch — say, sae, se HVff, (aai, saai, se) G, 

8f ge]> '^SSl' saith, says bob, tse]) 

(aez) Cp, 

895 weg twe55(e) way wei 

— wai iVbrfA.— away, awae HVff, waye 5W. (wai, waai, wee) G. 

wei-la-wei weilen wail weil 

Boeth, wellawell jEfc gr. generally walawa. 

Jnei tnag5 nay tnei 

— nai, mei Lay. nai AR, North, nay Ch — (nai) Sm, 



BIRST WOBD-LIST. 317 

4cL99Ja tdegenn die dai 

--de(i)2eii (deie) Lay. dei(3)en AR, ^egheiileghe adj^ dighe CM, 
die : lie vh MH, deje : felonye, deje : weye EBC, de : he, dy : I TM. 
de3e Harl, dy^eiyge AUP, dyiry^twysly And, deien PPl, die 
TKic7. dye : Emelyey deye : weye (;A--(dei) 8m, 

plefi^an plelen play plel 

plej^ $1, m Cp, Du, a Cp, Su. — pleien AB, fpl^j, tplawee North, 
tpbige &E, tplawen, tpleien Hv, fplei Ch, 

900 leger leir lair leer 

t)>6ira t)>e55re their Seer 

OEhi(^o)r%-'^ North, here CA— thein 7c/. (eei) &. (ee) Jn. (ee,«e) !.(/. 

fglan tegglexm ail eil 

|heil adj the^slenn hail heil 

— 'greet.' hal (ai) beo ^u 1 hailwur^^u! h»(i)l (hoi), hallen v& Lay. heil 
< hanus ' Prompt, hayle I heyle ! Wiel, hayl ! Ch. 

Beg(e)l sell sail sell 
(ai) a. 

905 megl sneil snail sneil 

{reiaa treg^Benn raise reia 

ttr^sta tristen trust trast 

U^uft in, trj^nit adf — (tristen, e) Lay. n Jnl. tmtte : wuste prt ROl. 
trosti Ay. traiste : Grist CM. tritte #6 : Criste MH. triste prt, traist 
adj'.^ysi vft jRBC tmst:hast, tryst :wyst prt, u Titf. u ^amI. 
tt, (i) PPZ. tristen, o vb, trist, o «6 Wiel. truste : ruste, triste : wiste, 
(e) CAr-trysU ck(; J3TV. (y) i56. 

regen tregjii rain rein 

— rein AR. ren Ay. — reyne, ay Td. (rain) O. 

gelegen pic tlejenn . lain lein 

— ^ileien AR — (ee) Cp — (leid) ty. 

910 ])egen 'pein thane t]>ein 

tsvein swain swain tswein 

OR swfin ' herdsman * — swein ' soldier * Ld. ai, ei Lay. swayn Ch, 

bleKen(e)8f blein(e) (ohill)blain blein 

^eimyija sf eimeri embers embea 

OE ftmyriaa pi — Prompt. aUo eimbre Prompt, 

Xpeim. "fpeggm them Sem 

him OE; late heom — hem Ch — (8em) O. (em) *em Jn — weak (ISom), (em). 

915 tsteiksf stelke steak steik 

— Prompt, — (ee) 8m, 

Iveik weik weak wijk 

OE wSo— wao 0, Lay. o AR. wa(y)k North, wook, wac : Isaao O 
wayke : layke vb Hv, wake : forsake TM, wayk, ey CA'(ee) 0, 

tsveigja sweien sway swei 

*bend.* 
tbeitosf beite bait belt 

tbeito tbejstenn bait beit 

'graie/'hant'—'panish.' baiten < feed ' CA. 



^ 



I 

I 

I 






18 



HISTORY OF ENGLISH SOUNDS. 



920 l^gde prt tle^d ptc laid leid 

]gde WS — laBide, le(a)ide Lay. ei AB. laid North, fUyde Ch — (ai) O. 

bregdan breiden braid breid 

fog tegge edge edj 

tfSS «^ egge egg eg 

OE Sg— «y Wiel., Ch. ey, egge Prompt. 



Jlfw 


leg 


leg 


leg 


— Econcan (legges) Lay. 






925 SfCg 


Beg(ge) 


sedge 


Bedj 


sl^cg 


slegge 


Bledge(liainnier) sledg 


wf eg sm 


wegge 


wedge 


wedj 


— wigge, e JPrompt. 








Jdr^gg sf 


dregges pi 


dregs 


dregs 


etan 


tetenn 


eat 


«t 


{ee)G. 








930 Betl 


Betel 


settle 


setl 


fe(o^tor 


feter 


fetter 


feter 


fretan 


teeten 


fret 


fret 


* devour.' 


« 






nftele 


netle 


nettle 


netl 


m^te 


tmdte 


meat 


mtit 


(ee) W. 








935 metan 


meten 


mete 


tmyt 


cftel 


ohetel 


kettle 


ketl 



— ohetil, k Prompt 01 k^til— r^ (kitl). 
get tj^t yet 



Jet 



e VP, Du, Ru. ie, i, y WS—get, iett, get Ld. jet {i^ Lay. jet All, 
AllP, And. tget GE. yet Ay. giet (ydt), giete : itte CM. fy^tto 
PC. 3itt, tyete MR. ^it Wicl. sat, jit PPl. yit TM—{i, e) 8m, 
(j«t) Mff. (it) Jn. 

(be)getan tbi^etenn get get 

e, eo VP, Du., Bu. ie, i eW8. y Z TFS^— beieton, bigetan Ld. gette North. 
bi3e(o)ten Lay. — (gjet) W. (git) Cp — vg (git). 

be-geten ptc tbijetenn begotten bigotn 

— bijeten Lay. beyete Ay. bygoten Wid. 



940 teter 






teter 


tetter 




teter 


b^tera 






tbettre 


better 




beter 


l^ttan 






tlettenn 


let 




tlet 


* hinder.* 














Sfttan 






tsettenn 


set 




set 


hw^ttan 






whetten 


whet 




whet 


946 ^f^ 






tnett 


net 




net 


bftst 






tbettst 


best 




best 


Bt^e 






tstMe 


stead 




sted 


aUo styde 


J>u., 


Bu. 


— etude Jul. 


stud AllP. 


fitede. 


■tide North. I, e 



CA— (insteed) G. (u) J». (e) Mg. (insUid) Beh. (inited) Sh, 



FIRST WORD-LIST., 319 

iBtmpyig tstidij steady stedi 

— 8t8B])6li Lay. u^fi. o/'stodefkat ' finn in one's place.' tee s^de. 

weder weder weather weCer 

e C*. 

950 medu mede mead mljd 

— ^meeth : heeth ( s hSi\f\ (mede) Ch ; 01 mJ9>. 

onedan ozieden knead x4fd 

(nheed) Q>. 

tredan ttredenn tread tred 

gebed sn tbMe bead btjd 

* prayer.' 

bedeoian beggen beg beg 

955 brfddan treddenn rid rid 

* reicae ' — anide imper. Jul, 01 ryj^ja ' olear away.* 

wfdd tweddenn vb wed wed 

'agreement.' 
bfdd tbedd bed bed 

reopan repen reap rijp 

VP. ripao W8. 

stf ppan steppen step step 

960 fbba ebbe ebb eb 

w^bb web web web 

nfbb neb nib nib 

' beak ' (neb) ' roiitnun * 8m. nib quite mod. 



dura dure door dor 

dura, dor 2>u.^lnre, o Lay,, North, u OS, Ay, o Wid,, Ch — (nu) Sm, 
(00, uu) Q. (duuer) iomeUmee Jm, (door) Ld, Bcht 8h. 

Airther flMVer 

futlier fhatSer 

— htfioT Z)i». ; anal, qf foie—fuiffor Lay., Alt, ibrther North, ferther. 
o Ch ; fikfther TJIf ; antU, of feoir — (furOer, fiirder, fitfOer) G, (Nrder) 
Cp. (fcrdir, fonlir) j9cA. (fwrlTor, fieerOer) £%— (leedar) ci^. 

965 ours tourrsenn ourse kees 
ooume vb Td, (u) G, 

tdorste durrste durst deest 

duiran tfi/*— doivte Ld, n (o) Lay, u -^Norih., GJS, fTM. dynte, i BBC. 
o (u) Ch, a by it^, of dunan. 

turf turf turf teef 

I 

— oIm torf. 

sourf seurf sourf skeef 

aUo Bcruf — aXto loorf, aorof. 

urnen ptc urnen run ran 

genumen ' coagulated ' — runnen : tunne AllB, yronne : sonne Ch, 



Air]x>r tfbrr]>err \ 



320 



HISTORY OF ENGLISH SOUNDS. 



]>iirh 



+]nirrh 



970 sptiman spumen sptim speen 

u, o W8 — also o. 

muman mumen mourn mon 

— u AS, -f North, u, (o) Lay. o Wiel, 00 (oa), mome : borne (»baiiie) 
Ch, ow rj»f— mo(u)me Td, (uu) BU. (uu) Wetc. («) Jn. (00) Ld. 

tumlan tturmenn turn teen 

also tyman — aiao ou, o, i, e. 

{through ]>raw 

thorough )>are 

\>\iT{h) WS. o VP, e, o Du, u Ru.—\>\iT(h), >uphe (prp) Ld. furh, 
o Lay. >ur3 (thoru) CM, pnT^, ))rych AUP. J>orw(e) PPL throj, 
thor^ And. thrugh TM—-itiOTow prp Td. ihrwch. prp Sb, (]>nix) O, 
(jn-Duh) BU, (J)uro) a«^ (J»roux) adv O, (froo) |wy Jn, 

fiirh Bf ftirh furrow fiurou 

— aUo furwe, fop(w)e. 

975 bur(u)g sf tburrh borough bara 

— burrgbeas pi 0. buruh (burh) AR. 

turtle tturrtle turtle(dove) teetl 

tbuU tbule buU bul 

ODan. ; OJboli. (u) Mg, Jn, Ld. 

wull(e) sf twulle wool wul 
woU Td. (w«l) Pr. (wtil) Cp. (w«l), better (ul) Jn. 

ftill tfUU ftOl tal 

ffwl ffVsf- (t») Cp. £d, Bch, 8h. (fi^kdm) fulsome ^<!A, Sh. 

980 ftillere ftillere ftiller ftder 

pullian puUen pull pul 

(u) Cp, irf. 

bulluc bullok bullock bulek 

ScifU, — alio balluk. 
wulf wulf 

— w(u)lf, (o) Lay. — (u, «) Cp, 
(u) 8h. 

hulo hulke 

'cottage,' 'ahip* — Prompt 

985 oulter eulter 

— aUo o — (ou) Cp. (koultar) Bch, (koulter) Sh. 

tbjjJM bolke bulk 

* sbip's cargo ' — Prompt, 

sculdor 8f (?) tshuUdre shoulder 

pi Bculdm -a, geecyldru — eculdre (soldre) Lay. ssoldren ROl, sehylderee, 
Bchulder AllP. sbulder, o Ch. u TM—{ou) Cp. (00) Ld. (ea) Bch. 
(00) 8h. 

]ni8 tJ^uBB thus Vas 

muBde muscle mussel masl 

990 tuso tuso tusk task 

tazaspZ IWS — tosch Prompt, alio tnscb. 

nist rust rust rast 

Q, hy anal, of dOst (?) — aho roast. 



wolf wulf 

(ulf) Jn, (wuulf ) EO, Beh. («) Ld. 



hulk 



co(u)lter 



halk 



koulter 



balk 



Joidder 



FIRST WOED-LIST, 



lust 



tlUBSt 



{ 



lust 

listless 

lyitan tb — n AS. i Xay., GE. o Aff. ou Aud. 
lyiUn. 

must must must 

' new wise * — u (o) Lai^, 



grust 
bustle 

love 



igust — 

995 tbustla bustelen 

luflan tlufbmi ) 

lufU tlufe j 

ly ttb SVg, loov C*. (uu) Sia, (a) G, 

Qn-bu&n abuven above 

ouffle ouflb oufT 

charter — aUo o. 



321 

last 

llstUs 
ii,\iCh,TM. i/ro» 

mast 

gast 
basl 

lav 



ebav 
kaf 



simoen pto 


sunken 


sunk 


sai)k 


1000 sorunoen ptc 


shxxmken 


shxxmk 


Jragk 


drtmoen ptc 


tdrunnkenn 


drunk(en) 


drai|k(e 


drunonian 


tdrunnknenn 


L drown 


draun 


— 4raxikxieii Wicl 


. drowne : towne TM. 




himgor 


thunngerr 


hunger 


hai|ger 


honger, anhonngred Td. 






hrung 


rong 


rung 


rai| 


1005 limgen sf 


lunge 


lung 


lag 


sungen ptc 


tsungenn 


sung 


sag 


^slunginn ptc 


slungen 


slung 


slai) 


swungen ptc 


swungen 


swung 


swai| 


stungen ptc 


tstungenn 


stung 


stag 


1 010 sprungen ptc 


tsprungenn 


sprung 


■praq 


wrungen ptc 


wrungen 


wrung 


rag 


oluBgen ptc 


tdungenn 


olung 


klag 


tunge 


ttunge 


tongue 


tag 


iongaTd. 








dung 


dung(e) 


dung 


dag 


1015 hunig 


thunig 


honey 


hani 


(huDl) G. 


{o)Bch. («)iSA. 






Jninor 


^under 


thunder 


]Muider 


— tlumer Pc, TM 


. Sunder QE. 






sunu 


tsune 


son 


san 


synn U Yg, 


{vL)8m,Q. {o)Bll, 




sounian 


tshunenn 


shun 


/an 


— scunien, Meoiiien (Miden) Lay, 

1 


■chones P«. 




stunian 


stunien 


stun 


Stan 


— aUo 00, 0. 








1020 gewunod ptc 


Iwuned 


wont 


twount 


(wiini) 0. 


(o)Bek. {xi)8L 
Y 







322 



HISTORY OF ENGLISH SOUNDS. 



1025 



1030 



1036 



munuo 


mon(e)k 


monk 


max|k 


Sonne 


tsunne 


eon 


san 


Bynn HVg. 








sponnen ptc 


Bponnen 


spun 


span 


gewunnen pic 


iwunnen 


won 


wan 


{n)8m, (0; 


) Bck, 8h. 






nnnne 


nonne 


nun 


nan 


oiinnan 


teunnenn 


cunning 


kanig 


kwning HVg. 






be-gtumen ptc 


tblgunnenn 


begun 


bigaa 


tonne 


tonne 


tun 


tan 


donn 


don 


dun 


dan 


huntian 


hontien 


hunt 


hant 


stunt 


tstnnnt 


stonted 


stantid 


< atupid.' stynton 


' blant, itupify.* 




* 


punt 


— 


punt 


pant 


onder 


tunnderr 


tinder 


ander 


hund 


thund 


hound 


haond ' 


hundred 


thonndredd 


hundred 


handred 


hand TPS^hoandred (u), hand Lcn^ 


f. handrat Jul, 


hnndr8t(h) Norih. 


— (hvndard) Cp 


etc. 






sund 


SlUld 


sound 


saund 


'swimming.' 








gesund a^j 


tsond 


sound 


saond 


sundor 


Bonder 


sonder 


Sander 


wondsf 


twonde 


wound 


wuwnd 



— ^w(a)nde (wonde) Xay. o North., Ay, on Ch — (oa) Sin, (au) 67. 
(aa) Ld. (eu) Bu, (nu) 8h. 



1040 



twonnderr wonder 



wander 



i*wondenn 

tftindenn 

■fgrund 



1045 



1050 



wondor 

wonden ptc 
ftinden ptc 
grand 

— granodwaU 0, — (on) &, 
grunden ptc gronden 
pond pund 

bonden ptc tbundenn 

sum tsumm 

^ — ME pi. synn (/br Bym) BVg. 

sumor sumer summer 

tsluma sm slumeren slumber 

genumen ptc fnumenn numb 

* taken.' 
ouman toumenn oome 

prt o5m — c6me gbH 0, cumm inflA. to com, came : dSm 2iH, come-: 
dSmerif. OJ«&«<ky&ma~-(u) a 



wotmd 

fbond 

ground 

groond 
poond 
bound 
some 



waund 

fitond 

graimd 

graond 
paond 
baond 
sam 

samer 

slamber 

nam 

kam 



PIB8T WORD-LIST. 



323 



orunil) 
flwuin 
onimple 
tumble 

diunb 



BWftm 

kratnpl 

tambl 

dftm 



toymlio oumelioh oomely kamli 

— oomU Prompt,— (kam^) O, 

enuna fonunme 

Bwiunmen pto swummen 

1055 OToxnp adj oroxnplen 

tumbian tumb(l)eii 

dumb tdumb 

— dom : boghtom PC. doumb (o) Ch — doinm(e) Td, domb Ck. 

pluoaUa pluoden pluok plak 

buooa tbuoo buck bak 

' hegoat *—ffen€raXljf bncke. 

1060 fdyhtig tduhbtij; doughty tdauti 

— duhti (o) Lay, dobty Ch, o, ou TM, anal, of do^an it^n, and dohte 
pri of dSab • avails '— (dooti; /». 

sugu Buwe BOW sau 

— rawe AB. loje Ap, lowe Ch. 

ftisol tuzel tawl fkul 

'bird* — ^faxel, fo^el trowel) Lay, fogbel PC, Tojal Ay. fow«l» foul 
Ch. fog], II OX— (fool) Sm. 

CfVigLe kuvele oowl 

— oole (korele) Lay, oowle Prompi. 
tugglig ugli ugly 

'fearftiL' 
1065 muog-wyrt mugwurt mugwort 

hnutu nute nut 

— nbote Ay. 11 AB, North,, OE, o Prompi. 

butera but(t)ere butter 

gutt gut gut 

buttue buttok buttook 

on \a baiinoM Umndary in chart. 

1070 rudig mdi ruddy 

wudu wude wood 

— wo(o)da CA— woode Td. («) Pr. (q) Op. (wisd), hotter (ud) Jfi. 
teudele — outtle(fl«h) katmj 

i>ii<a kuttelTiieh. 

uppaa prp tt^ponn upon apon 

Up on 'up on' — ^Qppon, nppen, up (uppa) Lay, opa Ay. apon: 
Joban Cm. upon, opon Barl, 

ouppe touppe oup kap 

u, o-^n Ld. o, tn CA— (Irabart) Ld; (kvpboocd) Mi; (k«bard) 8h 
oupiboard— (kabed). 



kaul 

agU 

magwaat 
nat 

batar 

gat 

batak 

radi 
wud 



1075 ]^yxaliaa 
' piarce. 



]>lrlen 



thrill 



T 2 



yrn 



3^4 HISTORY OF ENGLISH SOUNDS, 

■tjrlaxi tBtirenn stir steer 

wyxreet Bpl fwerrst worst weest 

y, ie, i Pait. y Anglian, e t^t— wur(e)Bt Jul. wont Ay,, Wiel. 
went Ch, 01 y^rsfc— (u) G, (uu, «) £0. (oo) B<A. («) i&L 

byr]>en sf bir]>en burden beedn 

— u Xay., AB, u, i, e TTie^. borden GE, Ch, birden feit.— burthen, 
dTd. (burdn) jBZ/. (b«E«en) Pr. (b«rd»n) /n. 

znyrjwaa i7nirr]>renn murder meeder 

— ^muz^ren Lof, mordren (oe, mor^eren) Ch. murder, o TM — ^mur- 
therere Td. (muifjer, murder) G. (d) Ld, 

loSo wyrsa cp twerr(8)e worse wees 

i, ie, y Pati. y VP, Or., Du., Ru. y IFi^— uuene, wserse Ld. warn 
North, w(u)r8e (o) Lay. o Ay. e Ch. 01 v^rri — (wuri) G. 

tyvB sm flrse ftirse faei 

— Wiel. fyzrii Prompt. 

)>7rstan t]>irrstenn thirst ]>eest 

]>urBt «6 — ^irrst, \iT\mt A 0. ^nnt (o) «& Lay. ^nrsten AM. >riiteii 
t JTor/A., t^ei4., TFt02. tluyBten, thrusty a/Si TM. }»urrt «& CA. 

iyrsta tfirrste first fi»est 

_ fryBt ^oitft. fyrst : brist tb : thrist TM. furst, forsi iitid. Tent Ay. 
first (e) CA. 

tbyrstsf bristel bristle brisl 

— brustles (i, y, e) Ch. 

1085 hymetu — hornet honit 

anal, of horn. 

oymel kimel kernel keenl 

— o^o u, e. 

wyrzn i'wiirmi worm weem 

— o Notih., AUP, TM, Ch. u Lay., AR. 1 Beit, wirm, wiim GJB— 
(uu) m. (u) G. (uu, «) ^0. («) JBdl, fi*. 

tyrkja irken irk eek 

' work.' 

wyroAn I wirrkenn work week 

i, y FP- y, i Dtt., Bu. weoro «5 — wiroen Ld. fwirk North, werken 
(7J9. w(u)rchen (e. 1) Lay. worohen AllP, Wid. werchen Ay. 
wirche : ohirche (e), werken (worehen) Ch. ii^. of weore. 

1090 my roe mirk(e) mirky (u) meeki 

— alio e. kfrom Scand. myrk. 
wjrrhta twrihhte wright rait 

— wurhte (wrohte) Lay. wurhte (wruhte) AB. wrighte Ch — wricht 
HVg. 

f^hto friht fright frait 

fryhte i?a.— oflruht adj Jul. frigt GS. 
wyrgan wirwen worry wari 

' strangle *— awurien AB. wirwln, worowen Prompt.— {vl) EO. («) jBM, 
Sh. 

myrg mirie merry meri 

myignis Cp—mjry North., AllP. mm:blri (-byrig) GE. myry, e 
Aud. murie (i) Lay. merye:berye, marie : Merourie, myrie:pyrie 

(» pear-tree) Ch. mery TM. 



PTEST WOBD-UST. 



325 



1095 myrg]> sf inirh]>e mixth mee]> 

— mur(h)'Se Lay. myr^e AllP. myrth TM. myrtha (a, e) Ch^{e, i) (?. 

be-byrgan tbirrgenn bury beri 

— ^borien Lay. birien OK biiyen, burien Wiel. u Ch. e Ay., Aud. — 
burieTVi. (j) Sb. 

byrgelB birieles burial beriol 

— biri(g)ele(8) OS. biriel, bnriel Wid. buriek Ch. bcrieles Ay. 

takyrta fif aohlrte ahirt Jaot 

— ■curte (seorte) Lay. i, • Vid. i Prompt, fe Ch. 

wyrt sf wurta wort woat 

* herb " — ^wnrte, o * wort.' wart, o, I, e * herb.' 

1 100 oyrtel tldrrtell 

— u Lay. e Ay. i Ch, 

hyrdel hirdel 

— also VL, e. 
wyrd sf wirda 

' fate ' — aUo u, e. werdee, ie Ch. 
gjrdei{B) tgirrdell 

— aiUo u, e— gerdell, gyrdle Td. 

gabyrd tbirda 



kirtla 


tkoea 


hurdla 


haadl 


waird adj 


wiad 


girdla 


gaedl 



birth 



bed]> 



—hxade (fi) AM. ^YOr^ North. UttitGE. hxaik,jAmd. barthe C%. 
also e. 

1 105 mylan Bf milne , mill mil 

— malne AB. mihie Wiel., TM. melle Ay. fmelle, 1 CAr— (milnar) 
Milner. 



oylen sf Idlna 

—aUo u— (kU) Ld. 



kiln 



kiln 



pyla 


pilwa 


pillow 


pilou 


hyU 


thill 


hill 


hU 


fqrUAf 


Billa 


Bill 


aU 


— also u. e Ch. 








mo fyHaik 


tflllann 


fill 


fll 


onyllan 


onullen 


knaU 


nel 


— alio y, i, e. 








^Igja 


— 


billow 


tbilou 


87lt 


tgillt 


guilt 


gilt 


(gwilt) Jn. 








gyldan 


tgildenn 


gild 


gild 


1115 byldan 


bildan 


build 


bild 



bylda m Qroiit^n, uy (i, ee) CK ey rif— byltt pH Td. ij Ck. 
(yy, ii, I, ei) Q. (i) Cp. (iu) /•. 

dyaig dial diaiy disi 

' fooliih ^-^iUto u, e. 

byaig biai buay bisi 

—^l$o u. biiy (e) CA^buiy Td. bnii Sh. (bicnei) G. (bisi, bUnei) Ld. 

oyaaan kiaaan kiaa kia 

tb 



listen 


lisn 


thrush 


>raj 


blush 

i) Sm, 0. 

Ii0t 


bUj 
tlist 


list 


tlist 


duster 


Idaster 


evil 


yvi 



326 HISTORY or ENGLISH SOUNDS, 

hlysnan Ustnen 
(lisa) •/», Ld, Beh, 8h, 

1 1 20 {nrysoe yruaohe 

— also ui. 

blyscan blusohen 

* rutilare' Aldhgl. — u Ck. also i — (u) Sm, O, 

gelystan tUsstenn 

* desire ' — -fe, fi, u Ch. 

hlystan tUsstenn 

'liBteD.' 

clyster cluster 

— also o. 

1125 yfel tifell 

_uyel (heyele) Lay. iuil (il) CM. i GE, Wicl, e Kf, AUP, And, 
e C*— yveU, evyU Td, (iivl) G. (iivU) Bt, 

]yft left left left 

left Mnanii' MoGl. lyftSdl ' parftlysif '^a, eo, i Lay. e Ay., fCM, 
fBBC, CA— lyfte Td. 

XkyttA tlefltenn lift lift 

— e, i CM. ly fte : thryfte TM. 

geolyfte clift oleft kleft 

'* Bectilem ' Bogl. — i Ch. u Prompt. 

ynoe inohe inch inj 

— also a I e — e Td. 

1 130 Jiynoan t]>innkenn think ])ii)h 

' seem.' |?^ncan ' think * — ' think ' expr. hy ^ennkenn 0. ; ]>enken "^RBCf 
Wiol. ; ])enchen AB, Ki ; he |)eng)) Ay. * seem ' saepr. hy >innkenn 
0., GE; >anchen AR ; >anchen, 1 Lay.; I^ingken, me >ing> Kt. )ane 
' think/ * seem ' CM. fthencheni thenkJen, f^ithynken * think, think 
of;' thinken (e) 'seem' Ch* 

lynes line llnoh(pin) linj 

myne menow minnow minou 

mynet mint mint mint 

' ooin *^^also munet, menet. 

oyning thing king kix| 

oynin^:, qyng firom hsginninff, the latter the nnemphatie form — kyng, 
king Ld. king Lay,, AB, Kt—i HVg, 8b. 

1 135 dyne dine din din 

*bryne-st&n brimston brimstone tbrimsten 

' barning-stone * — also u, e. 

])ynne pinne thin ))in 

—i, te CA—thynn Sb. 
synn sf tsinne sin sin 

y HVg, 8b. 

oynn t(mann)kinn kin tkin 

— mankinde hy anal, o/gecynd. 

I J 40 mynster sn tminnstre minster tmlnster 



•t^^ten 


tstinntenn 


•tint 


* blunt*' • rtupify.' 






dynt 


tdinnt 


dint 


'ttrok«'— teC4. 






gemynd efn 


tminde 

• 


mind 


geoynd sf 


tkinde 


kind 


geoynde adj 


kJxLde 


Und 


' natural.' 






1 145 tynder 


tinder 


tindo] 



FIRST WOBD-LIST. 327 

•tint 

dint 

maind 

kaind 

kaind 

tinder 

— i Lay, u \B€9t,, Prompt. ; 01 tundr. tendrin * bum ' intr Jul, ; 
a»aL of oat^dan — («) barbate Cp, 

tryndel trendlen vb trundle trandl 

alio 6, 86 — (tnnd) Jn, 

bsrndelle bundel bundle bendl 

— Prompt,, Wid,; anal, of gebundaa. 

hymlio humlok hemlock hemlok 

hymblice 'eicuta' Ep., hymlioe Cp, 7, e Leeekd^ — humlock Prompt, 
aUo heiiMluo. 

trymman trumen trim trim 

* oonfirm.' prt tiymede— Mtrfy and raro, 

1 1 50 brymme em brimme brim brim 

In IW8 confused with brim * ocean/ toAicA was orig. netU, 

tmyk 8f muk muck mak 

oycens kiohene kitchen kitj^ 

-Hsnchene Lay, u AB^ e Ay, 

bryoe breche breach br^tf 

— u AR, aleo i. e Kt, or i^fi. ^brecan. 

oryco sf (9) crucohe erutoh kratj 

1155 tyxensf flxene Tixen vikan 

fixenhyd Leeekd, 

flyhtsm fliht flight Halt 

'flying'— vluhtilB. 

flyhtsm -fflihht flight flalt 

'fl«ebg'— fluhtiiB. 

ryge rie zye rai 

— aUo ru^e. reye : prey tb (rie) Ck. 

hrycg rig ridge ridg 

— rug (rugge) Lay, rig : big aty He, reg Ay, 

1 160 myeg migge midge midg 

brycg 8f brigge bridge bridg 

— brugge Lay, brigge : Oaniebrigge» brigge : oollegge CA-*-(bred5) Sm, 

{X)Q. {e)Jn, 

•oytel sohitel shuttle Jatl 

—aUo e, 

•oyttan eohetten aliut /at 

— e Ay , +CA. often u. rarely i — icfait Ck, . 



328 HISTORY OF ENGLISH SOUNDS. 

onyttan knitten knit nit 

1 165 grytt grutten adj grit grit 

' coarse meal ' — gmUeae bread AR, 



pytt 


pit 


pit 


pit 


— also u, e. 








Jfly«a 


tflittenn 


flft 


flit 


dyde 


tdide 


did. 


did 


— dude (e) Lay, 


e Kt i (e) CA— (dad) harbare Cp, 




Boydd sf (f) 


schudde 


shed 


Jed 


Hudelingsoydd chart. — Prompt, 






I I 70 hype 


hipe 


hip 


hip 


orypel 


oripel 


cripple 


kripl 


— also u, e. 








oH-slyppe 


ouflloppe 


oo'wslip 


kauslip 


olyppan 


olippen 


dip 


klip 


' embrace.' 








dyppan 


tdippenn 


dip 


dip 



deppetende (=(b) Cp. depa vel dyppe Ru, d7p(p)an IW8 — alto a, e. 
— depe Td, 

1 175 tfltybb stable stubble stabl 

' stump ' — stobul, stubbil Prompt, stubbe ON. 



• O. 

l6r-loren forloren fbrlom felon 
(forlorn) Q, (00) Ld, 

sworen ptc sworen sworn swon 
(swoom) 0, (suum) Cp, (soom) Jn, (soom, swoom) Ld, 

sooru so(h)ore score skoer 
landscoru chart, — sohore Ch, 

sooren ptc tshorenn shorn Jon 
(uu) BO, (00) Ld, (o) Bch. (00) 8h. 

1 1 80 spora spure spur speer 

o, u Wgl, — spure (o) Lay. spure : dure ON, o CA, Prompt, 
woruld sf twe(o)relld world weeld 

eo VPt Ru, eo, o W8, o 2>i*. — ^weor(e)ld (worl-) Lay, fwerd, warld 
North, ww{\)dGE. fwerd RBC, warld TJIf. werd, wor(l)d : lord 
Aud. werde Prompt, world, wordle Kt — (world) G, (wvrid) Pr. 
(wwli) • worldly ' Jn. (uu. «) EO. («) Bch, 8h. 

for fforr for fber 

(for) G. (forget) etc 0. 

beforan tbiforenn befbre bifber 

—also biforr >att 0.~(bifoor) G, (foor) fore Bt. (foortel) foretell G, 
(foorward) bU, (forord) forward Jn — (fored) forward v^, 

tfroren ptc ftoren frosen frouan 

%^/lqf freonsi, 

1 185 tgecoren tchosenn chosen tjousn 

— if^fl. of cSosan. 



FIE8T WOBD-LIST. 329 

gor 8n gore gore goer 

'dang.' 

toren pto toren torn ton 
toorn Ck. (ianrn) EO, (o) Beh. (00) Ld, Sh. 

boren ptc tborenn bom(e) bon 

(00) < natns/ (o) < MUtus ' Bll. (o) G (teithout dUiingvJUking 

• borne *). (uu) ' UjuUtut ' Cp. (00) ' port^ * Mg, (o) * pwrturitui ' 
Cp. (00) 'n<* Jf^. (00) * natiM,' (00) 'l»tu»' Xd. (uu) E0\ (00) 
£% borne, (o) Bc/^; (00) /S% bora. 

borian borien bore bor 

1 1 90 fbr)) tfbrr)^ fbrth ft>]» 

(faur» (?, Cp, -ffO. (00) jW, Boh, 8h. 

gef6r)>ian iforj^n afford elbd 

' forward '— ' perform.' laier afor>en ' proWde '— (afuurd) Bt 
noTp "fnoTrp north no]^ 

mor)x>r sn mor]>er murder meeder 

moTp im W8 prote, myr^ran vb — ^morO (mor^re) Lay, murther North. 
morthyr, morther Aud, ufrom myr>ran. 

hors thorrs horee hos 

(o) 8m, 

119S fforet gorst goree gos 

borsten ptc borsten burst beeet 

— Laiy, broeten Ch — u from prt pi bonton — oy (bast) . 

bom horn horn hon 

(horaed) O, (boom) Cp, 

)x>m t^rm thorn )xm 

ooC*. (oo)Bff. 

com tcom com kon 

ooorne Td, (00) Bll. (o) G. (o) Bch, (00) 8h, 

1200 storm storm storm stom 

fbrma "Hbrrme fbrmer ftnnor 

— nvniere WuH, 

ii^rmesl fbrmeet foremost fomest 

— Lay, aUo u, i. formast CM, 

store stork stork stok 

forca forke fork ft>k 

1 205 geworht ptc twrohht wrought rot 

— WTohi Ld. iwraht Jtd, tmn^, twroji prt AUP, fwro^ CM, 
fwrogi OS. wro(u)ght CA— wrogbi Td. (root) Pr, J% EO, (o) 
Bch. (00) Ld, 8h. 

sorg tserrjhe sorrow sorou 

■oeigendi JS^.— eenrsbenn vb 0. lorje, eeorwe^ eeorbful Lay. seorawe 
(sorbe) AB. lorje Ay. Mru CM. eonre CA--(aoro, waroou, Mrooui) O, 

morgen morwe(n) morrow morou 

o Ep. a VP. o, e WS, o, on merne Du, o, m, meigenne, mame Bu, 
— ^to morjen, a, m (morwe) Lay. morjen Ay. morwe Ch. 

*morgenung morwening morning monig 
(00) Jjd. 



330 HISTORY OF ENGLISH SOUNDS. 

borglaa borwen borrow borou 

(baroou) 0, (boru) Pr. (booroo, boroo) /». 

I2IO ort-geard oroliard orohard otjod 

ore-, ortr PtuU oregyrd, ordoeard, orce(a)rd IW8 — orcherd Lay. 
Boort tshorrt abort Jot 
(o) flt. (o) Beh, (00) 8k. 

port port port pot 

(oo) Ld, 

or-dfil ordal ordeal odJljl 

ordol LatOB — Ch — %i^. qfdsl. 
bord tbord hoard hod 

— hoord : woid C*— -(uu) Pr. (hwd) hord EO. (oo) Ld, Bch, 8h. 

1 2 15 word tword word wood 

— o, eo, u Lay. o, u (?.©— (uu, u) Bt (u, o) G. («) Jn. (uu, «) EO. 
(«) Ld, Beh, 8h. 

fbrd fbrd ford fbd 

—aUo for|> — (uu) Jn. (oo) Ld, Sh. (o) Bch. 
bord tbord board bod 

—bord : word CA— bourde Td. (uu) £i. (00) G^. 

Jskorpna tsoorronenn scorch akotj 

* ahrivel up ' — also scorchen. 

houl 

hoU 
)x>iil 

Joul 
Btouln 
foul 
koul 

« dal 

' fooliah '— u Eath., Prompt., Ch. i TM. o rare. 01 dul ' oonceit/ 

tbol bole bole tboul 

' trunk of tree.' 

Bwollen ptc swollen swollen swouln 

(Boobi) /n. 

onoll knol knoll tnoul 

(nhoul) Cp. (nool, bouI) EO. (nol) 8h. 

1230 toll tol toll toul 

aUo toln (oou) Sm. (00, ou) W. (00, ou) EO. (ou) Beh. (00) 8k. 

boUa bolle bowl boul 

(oou) a. (ou) W. (ou) Cp. (oou)/». (ou)^O. (ou)BcA. (00) fi^*. 

bolster bolster bolster boulster 

(00) Ld. 

woloen an welkne welkin twelkin 

'cloud' — BO wolcne Ld. w(e)olcne pi Lay. welkyn AUP. walkne 
GE. welkne Ch. 



hoi an 


hol(e) 


hole 


— ^hol Lay. 


hole CA— (00) 8m. 




1220 hola(g)n 


hoU 


holly 


)k>1 


jK>l(le) 


thole 


—Prompt. 


' cUviculA.* 




scolu 


— 


shoal 


stolen ptc 


stolen 


stolen 


fola 


fole 


foal 


1225 col an 


col(e) 


Qoal 


tdol 


tdill 


dull 



FIBST WOED-LIST. 331 

folo tfbUo IbUc tfouk 

(foolk, look) (?. (fook)/i». (tok) BelL (took) 8h. 

1235 holh sb holwe adj hoUow holou 

< hole '— holh aty Lay, holou Prompi, bolgh TM, holwe (holewh) CA. 

fblgian tfoUghenn fbllow fblou 

alio fylgan — foUh imper, 0, folien, u (fol^en) Lap. aolewen (folhen) 
AB, folwen CA^folooa) G, (folou) Pr, Jn, (feobo, foloo) com, Jn, 



molten ptc molten 


molten 


tmoultn 


oolt an oolt 


oolt 


koult 


{o)Bck. {00) 8h, 






bolt) sn bolt 


bolt 


boult 


(00) W. 






1240 soolde t8(h)o]lde 


should 


Jud 



— floolde, u Ld, iciilde. (solde) Lay. rchulde AH^ AUP, Wiel. ■o(l)d 
North, achold Aud, o BQl, Kt. u hy inji. ofplprt eoulon. fo (u) 
Ch. to, u, a TM. B weak ; ep aoeal^Bhulde Td. (Juuld) (7. (Jould) 
Pr. (Juuld) Cp. (Juud) /», ^c*. (Juuld, Juud, u) Ld. (Jud) 6h. 

woULe twoUde would wud 

a rP, 2>tt. o, a Att.— o, a, (o) X^. . o, (a) ill?, fwald, i North, u, o 
GR wolde : were f olde ' filled ' Uv. u Be$t. fi, a, wolde : hulde BBC. 
o CA— wowld, wld HVff. wold(e) Td. (wuuld) J//, (would) Pr. 
(wuuld) Cp. (wuud) Jn. (widit, woudst) harbare Cp ; (wuust) Jn 
wouldBi. (wuuld, wuud, u) Ld% (wuud) Beh. (uuld) Fr. (wud) 8h. 

molde molde mould mould 

(00) Ld. 

gold tgold gold gould 

gQld(e) Td. (goould) G, (00, ou) W. (ou) Pr. (uu) Jn. (uu) 

Bu., Sh. (guulimi» «7i». 



ifroptk trope 


ftoth 


trop 


(fro» (?. 




■ 


1245 broJ^SQ hrclp 


broth 


bro)^ 


(o)Jfc*. (00) fl*. 






moppe mo]>(]>)6 


moth 


mop 


moh^ !>«., Bu. — ^mou3)«, mou^te 


WieL 




hose hose 


hose 


houfl 


rose rose 


rose 


roua 


geloeod ilosed 


lost 


lost 



(00) Cp. (o)B€k,8k. 

1250 nosu nose nose nous 

— neoee (o) Lay. nese North., fTJIf. fnoie, e Best, nase Kt. 

ttmrn-pfrei mcmepirl nostril nostril 

miayytlk, uteweard noeterle late — ^nosethirl ('^) Ch. neeetlurl Prompt. 

tmoel mos moss mos 

OEuOoe. 

drosne dros(e)ne dross dros 

'leee.* 
I^kross en oros cross kros 

— Lay. 



332 



HISTORY OP ENGLISH SOUNDS. 



poust 



tou 



of 



toflE(e) 



1255 fro0t tfrosBt troBt frost 
(o) a, (00) Cp, (o) Bch, Sh. 

post po0t post 

(uu) EO, (o) Beh, (00) 8k, 

*tow tow tow 

towlic wecxro ' textrinum ' Wgl. 

( o^ o' OV, e(v) 

(off Of 

prp and adn — offe adv 0. of (o) prp CM — oflf prp HVg. of adv Td. 
(ov, of) prp G, (of) prp W. (0) prp Ld, (ov) prp Bch, Sh, (of) 
advSh, 

ofto foferr, ofrr over 
(o)BU,0. (oor)/n. 

1260 ofen sm toffne oven 

— eUewhere oven. 

*hofeI hovel hovel 
(htnrel) EO. (hovl) Bch, (hovU) Sh. 

Boofel sf schovele shovel 

(/aul)Bfl. (J9ul)/n. 



ouver 



avn 



hovl 



Javl 



tsoo&ttan 


schoven 


shove 


Jav 


■tofe 


— 


stove 


stouv 


8to& ' balnaum * 


Cp—ffom Dutch Btove (!). 




1265 tgewefenptc 


woven 


woven 


wouvn 


coflt 


oove 


oove 


kouv 


•ohAmber*— C3f. 






clofii 


olove 


clove 


klouv 


duf-y clof-. 








olofen ptc 


tclofenn 


cloven 


klouvn 


ibxes-glofii 


foxesglove 


foxglove 


foxglav 


1270 *of-fiEai 


offal 


offal 


ofl 



01 ofl&ai ' dimimition ' (ofal) G, 

oft tofPte often ofh 

—often GE, fofite, often CA— (often) G. (oofn) Ld. 

tloft sn to lofft adv loft loft 

*ur.' & loft' up.' OiSflyfi— lifft'air'O. loft < aolAriam ' Prompt, 
oroft oroft croft tkroft 

on tonn(e), o on on 

— on, o9eXd. ofn) Jul,, Ch. ane.iQ. 01 tL 

1275 )K>nne t]>an(n)e then 8en 

o, a Ru, m IWS — ^anne Ld, ^nne, e, (a), |>ane^ yon. Lay. ^eonne 
(e) AR. )wme Kt, f^an North. 6aii(ne) GE, ^enne JllP, fthen 
Aud, thanne, fthan, tthenne Ch—ddeu HVg. then Td. 

jK>nne t]>an(n) than Sen 

0on VP. Oon(ne) Du. o, a Ru. — ^)»an(ne) Ld. ]»aa(n)e, ))en(n> 
Lay. |>en(n)e, Jwn AR, )>aD(n)e Kt. )»n North, 0an(ne), t^ne 
GE. ))en AllF, then Aud. than Ch^e Td. (e) G, 



PIEST W0BD-LI8T. 



333 



hwonne twlia2m(e) when when 

hao(e)iine, boenne Du. o, a, • Mu. m IW8 — w(h)oime, whaenne, 
w(h)0niie (wane, wan) Lay. hwon(ne) (hwen) Alt, huanna Ay, 
quen, whan North. quan(n)e, quan Q-E. fwhenne RBC. quen 
AUP, when, whan And. whan, (e) Ch — (wen) better (when) /n. 
(i) Mg, 
loo sn lok look loh 

■ooiaa sokyn soak souk 



Bokyn 

1280 aonooian amokien smoke 

(00) 8m. (it smuaks) Sm{ ^ 8m8ooe». 



amouk 



tgeeprecen ptc 


apoken 


apoken 


apoukn 


— Bpeken rare. 








oeooian 


ohoken 


choke 


tjoiik 


— alio chekin. 








onooian 


knok(k)en 


knock 


nok 


n. W8. 








geoo sn 


tjooo 


yoke 


Jouk 


1285 Jpoki 


poke 


poke 


tpouk 


'bag.* O^pohha. 






brooenptc 


broken 


broken 


broukn 


hooo 


hoo 


(holly)hock 


holihok 


rooo 


rokke 


rock 


rok 


stanrocca (aoopulc 


iniro) Atdhgl— 


aUo roche,/r Fr. 




looo 


lok 


lock 


lok 


'lock of hair.' 








1290 8000 


aok 


aock 


aok 


amoo 


amok 


amock 


amok 


BtOOO 


atok 


atock 


atok 


floec 


tflooo 


flook 


flok 


0000 


00k 


cook 


kok 


1295 ooooel 


ookkel 


(oom)cockle 


kokl 


oroooa 


orokke 


crockery) 


krokeri 


oloooian 


olokkln 


cluck 


klak 



dock 

pox 

cough 



dok 

poka 

kof 



MO wamb cloooe> Zeeehd, 

doooe dokke 

poocaa pi pokkea 

1300 oohhettan co(u)ghen 

(koouh) 8m. (kof) W. (koof ) Mg, Ld. (kof) Bch, Sh. 

oxa toxe ox oka 

fbx tfox f6x foka 

box box box boka 

'box/'box-tree.' 

dohtor tdohhterr daughter doter 

— dohter, douter Xay. doubter (dohter) ili?. dojterjy. dogbter ^or^A., 
TM. doaghter Ch. aUo au.— do(a)ghter Td. Tdooxter) 0, (dafWr) 
wmetimn Bt. (dooflnr) oee, Jn, (mb) Ld — tg (daator). 






334 



HISTOEY or BNGLISH SOUNDS. 



bought bot 

(boot) Cp, (boot, boot, boft) /n. 



togian 



1305 bohteprt tbohhte 

— ^boagbte CA^boxt, boouzt) C?. 
(o) Beh. (00) Ld, 8h. 

Xhrogn. sn pi roun 

flogen ptc flowen 

(oou) O. 

rtojen ) 
ItoggenJ 
*wan-togen ptc wantowen 

wantones Td, (woonton) Bch. (wonton) 8h. 

1 3 10 trog tro(u)gh trough trof 

— trou(gh) Prompt,-^iror) W. (troo) Jn. (troof) Ld, (trof) EO, 
Beh, 8h, 

boga bowe bow boa 
bo SVff, 8b. rayneboll Td. (boou) 8m, G. (boo) Cp, Ld^ 

frogga fix>gge 

late; alioayt gg. 



roe 
flown 

tug 
wanton 



rou 
floun 

tag 
wontan 



dogge 



frog 
dog 



dooga 
Bos^l. 
otor 
1315 rotian 
hlot an 
])rote 
gesnot 
Bcoten ptc 
gesoot sn 

1320 flot 

on flot, * afloat.' 

flotian 

very late, 
floterian 
mot sn 
oot an 

1325 grot sn 

' fragment ' (00) EO, Ld, 

tpotian tbuttenn 

JBfeU — aUo pa(t)ten, poten. Fr boier. 

botm botme bottom 

— aho hopem, 
splott 

dott- 
'niMsa.* 

1330 onotta 

(nhot) Ld. 

dott — 

'heiAo{hoiV Leechd. 



oter 

trotenn 

tlott 

])rote 

snot 

ahoten 

ahote 

o flote 

flotten 

floteren 
mot 
cot 
grot 



otter 

rot 

lot 

throat 

snot 

shot 

shot 

(a)float 

float 

flutter 
mote 
oot 
groat 

but 



8p(l)ot 

clotte 
knotte 



spot 
(dot 

knot 

dot 



frog 
dog 



oter 

rot 

lot 

])rout 

enot 

Jot 

Jot 

flout 

flout 

flater 
mout 
hot 

tgrout 

bat 

botem 

spot 
hlot 

not 

dot 



FIBST WOBD-r.IST. 



335 



ne plot ne ploh L^eehd. 

Boden pic flod«n 

god tgodd 

»^efi. Goddes Ch. 

'335 god-Bpell tgoddapell 

^g(Mpel AB, Ch. 

trod6n ptc troden 

bodian bodien 

bodig tbodij 

toddi odde a4j 

' triAOgle/ ' odd niunber/ 

1340 oodd ood 

open toi>6(nn 

hopa hope 

oopor ooper 

drop* drope 

--droppen th North, etc. 

1345 Popig pop! 

hoppian hoppen 

loppeatre lopster(6) 

■oppiaa Boppe Bb 

■toppian atoppen 

1350 atxop — 

vd arwi>^ (itnippos) Wgl, 

ftttor-coppe ooppe 

'Bpider.* 

oropp crop 

* duster/ 

ijtooppa oroppea 

•piok/'gnuw'— iliZ. 

topp top 

• ■ammiV ' top * ( ■■ playthiDg). 



{ 



plot 
blot 

flodde]i(ed) 
god 

goiq^el 

trodden 
bode 
body 
odd 

ood 

open 

hope 

copper 

drop 

poppy 
hop 
lobster 
sop 
atop 

(raaor)Btrop 
strap 

oob(web) 

crop 

crop 

top 



plot 
blot 

sodn(d) 
god 

gO(q;»l 

trodn 
boud 
bodi 
od 

kod 

oupn 

houp 

ooper 

drop 

pop! 

hop 

lobster 

sop 

stop 

strop 

strsp 

kobweb 

krop 

krop 

top 



ft. 



1355 rft Bf ro 

^Jnrft ]yro 

' straggle^' ' obftinacy.' 

sl&sf Slo 

swft t 



roe 




ron 


throe 




]nrou 


sloe 




slou 


so 




sou 


Ld, wluirM 


§U0 


. ■(w)». 



M(M>} 



336 HISTOEY OF ENGLISH SOUNDS. 

Lay. Bwa, so Jul. s(w)o Kt, 8(w)ft CM, sws, m PC •(w)o 
Btf««. 90 AllP, Ch^-^ ffVg. 

wft twa woe twou 

— ^woa (wo) La^, wumme (->w& mS) Jul, 

1360 hwft twha who huw 

— hwoft (hwa) ilJB. qua North, quo (?J!?. wo J9«t^. (w)ho BBC, PPL 
huo Jy. who Ch—hw HVg. (whuu) JB/J, 0, Pr. (huu) Cjp, Jn. 

fft 16 fbe tfbu 

tfrft tfr* firo frou 

— frtkprp Ld. 

n& fna no nou 

no VP, Du.f Bu, no, na eWS, na 2TF)S^— na Zd. na, neo, nea, no Lap, 
no Jul. no, noa^AB. nummore (-snft mbe) Beff, no TTicl. nathelee. 
(->n&))yUn) CA. 

gftn tgaa go gou 

alm> gan|i;an — also ganngenn 0, gan, gon, 5(e)ongon Lay, — (go) O, 
(goo, guu) W. 

1365 t& sf to toe tou 

twft ttwa two tuw 

neut. and fern. — twa men Ld. t(w)o GE. to (too) BBC. to Prompt. — 
(twuu) BU. (tuu) O etc, Cp. (tvpins) Mg ; (tvpena) Jn ; (tvppias) 
Bch ; (tvpena) Sh twopence — (tapons). 

dft sf do doe dou 

ftr sf ore oar or 

hftr hor hoar hor 

1370 h*r-htoe horehuna {J^} hound hor(h)a«nd 

rftrlan roren roar ror 

Iftr sf flare lore lor 

8&r Bor Bore sor 

B&rig taarij sorry sort 

a, o Paet—uii, m (0) Lay, seri, o OE, loory Ck, ofrom warg. 

1375 mftre tmare more mor 

— oa AB—<m Td. (00) 8m, G. (mooor) EO, (moor) 8h, 

gftra gore gore gor 

*oomer* etc, 

gftr-ldao garleek garliok gaalik 

bftr bor boar bor 

(buur) Cp. 

hftl thai 1^^^^ ^^^ 

(hale heU 

— fhale Ch; from North.— hoUomo, whole Td, hoole Ck, (whool) 
Bull, G, (hoolflum) wholeMme G. (haal) hale G. (hool) W., Jn, 
(hool, whool), (holi) wholly Ld, (whool) B<A, Fr, (hool) Sh, 

1380 hfilig thalij holy hotiU 

— pZ hall^he 0, 



PIBST WOBD-LIST. 337 

hfiligdAg haUdai hoUday holidi 

— holiday (o) Ch, baleday Aud. 
h&lgian fhalljhenn hallow httlou 

mSl mol mole moul 

'mark.* 'stain.' 
gedftl an tdale dole doul 

tSgedSl ' dUtribotio ' — dole /em. in Lay. idol * separation.' 

1385 pftl pol pole poul 

ft]) fa)) oath ou]) 

1ft]) tla]) loath tlou]> 
(lo]>) Bll (lojsum) loathsome 0. (lo» Ld, (o) Bch. (00) 8h, 

Iftpian lo])ien loathe loutS 
(loo«)BW.JW. {00) Bch, 8k, 

wrft]) wra]) wroth fro]) 

—00 Wid., CA^-oo Td, (o) BU. (00) <7.— (o) yrom wrath wbtt, 

1390 olft]) tela]) doth Uo]) 

(o) a. (o) Beh. (00) /SA. 

clA]>a8 pi tola]>eu clothes ]dou(t$)s 

— folosejpl r3r->(ldoo«ed) ptc Q. (kloos) ^«/i, £%. 

:tbft})lr tba]>e both bou^ 

OJ^ hft, bSgen, bfttwS — ^bat^e, beien Ld, alto ba, be^^enn 0. b(e)o8e, 
botwo il fi— booth Ck. (bo» &. 

hfts hoB hoarse hos 

— ^hooB, hon Prompt., Wiel. — (hoors) 8m — the (r) U imitative, at in 
Dutch heeracb. 

trfts ras race reis 

OE rfts — n, e, ea, e Lay, ras North, — a^Vom North. 

^395 ft-rfts trfts (a)rose (e)roua 

])fts pi ]>os those t$ous 

' these'— >a < those/ ^ise 'these* 0. 

tmftse mose (tit)inoiise titmaus 

ftsoian tasskenn ask aask 

so, hs, z W8, hs, xs, so Bu. — axen Ld, azien, m Lay, aakien, axien 
(easkien) AR, esse, prt esste RQl aishest ON. askede K8. oksi, 
aksi ily. aske iNorth., QE, Aud. faxe (sk) CA— aish BVg, axe 7d. 
afdL €t aks iSni. (s) Bch, 8h—^g (sbx). 

Iftst last last laast 

aUo SB. ' track ' — alto e. 

1400 gftst tgast ghost gOTist 

(00) Cp. (gooeli) « ghostly * Jn. (uu) JFO. (00) Bch, Sk. 



a. 

ft-wlht tohht aught tot 

owiht Bu. awnht, auht, aht W8—aUo awihht 0. aCwi)ht, oht Lay. 
owiht, out iiiS. io^i North., TM, o^t AUP. t^t BGl, Ay. t^ughi, 
onght Ch. 

Z 



slftw 


tslaw 


Blow 


8lo SVg, 






8X1&W 


Bnou 


snow 


— snaw Ay, 






1 410 nft-wiht 


tnawlhht 


naught 


no-wiht 


tnohht 


nx>t 



338 HISTORY OF ENGLISH SOUNDS. 

rftw sf rowe row rou 

a, m o a, e, (a\ AB. fraw North. rawe:ow0 vb MIP, o Prompt, 

■feCh. 

lAweroe larke lark laak 

— Ch, laueroo Harl. laverok Gotcer, 

jf&WBXL thowin thaw po 

—Prompt— Kw Kt (?). 

1405 ])r&waa }>rowen throw ))rou 

* twiet ' — ^^we Ay. 
Bftwan tsawenn sow sou 

— zawe Ay, — (aoou) G. 
sftwol sf tsawle soul soul 

— zaule Ay, — bo(w)1 RVg, (oou) Q, (ou, 00) W, (00) Pr, (ou) Cp. 
(oou) Jn, (00) Bch, Sh, 

slou 
snou 

'not 

not 

iio(wi)ht VP, na(wa)ht^ noht P€ut. iio(wi)hiy nenilit Du, — ^nohi^ a 
Ld. iia(wi)hty nawit, nobt Lay, nowihi, nowt (nawt) AR, no^t 
BGl, nocbt, a K8. nagt Ay, nojt CM, fnogbt, fnott TM, fnoht 
GE. no^t, fnot AUP. tnoht, fa ffarl, no(3)t TTiel, fnog^i (ou), 
fnaught, nat (o) Ck. — ^nought! * naughty' Ck, nott Td, (naozt) G» 
(nooft) oceanonaliy Jn, (noot) Pr, (00) Ld, 

mftwan mowen mow mou 

— mawe Ay, 

orftwan orowen orow krou 

— crawe : mawe ( ■■ maga) Sari. 

or&we crowe crow krou 

141 5 on&wan tonawenn know nou 

— knAwe Ay. CDKwe AUP. kaow : Bchewe Aud. — kno HVg, (knobu), 
(knoouD, knoon) ptc G, (knou, 00) W, (nhoo) Cp, Ld. 

n, ( knouleohen acknowledge eknolidj 

( knouleche knowledge nolidj 

— aUo knoulage — ^kQo(w)le(d)ge ah, tb Td, (knooaledj) O. (hnoledj) 
Ld, 

bl&wan blowen blow blou 

— blawe Ay. 

iL'hwmper to(}>e)rr cj or or 

a(w)])or — also oJ»J>r 0, oww)>err pm 0. o(u))>er ej Ld. oJ>er ef 
Lay., ARy Ay. ou|7er CM. or CM. ou]>er (ei)>er) . . or 'eitiier . . 
or' Ch. 

n&-hwflB}>er tnoww]>err nor nor 

nobw8B|9er, noujier, na(w))>ere7P'i$ — ^nou))er Ld., Lay., AR, no(a)^r, 
nauJMBr North, no(w))7er, nawder TM. naw]>er AllP, noujwr Aud, 
nor PPl, Ch — ne(the)r Td — we nsegper. 

1420 hl&f tiaf loaf louf 



FIB8T WOfiD-LIST. 



339 



hlAibrd tla&rrd lord lod 

—lauerd CM, laaerd» loueid, lord PC, ilcacA MIP, Aud. lo(Qe)rd 
KS. Ihord Ay. ]o(o)vd Ch^oo) 8m. (o) G. 

hlAf-nuBflse Iwinnianie lammM tlsmai 

— Ld, lammease Prompt* 



olftfire 


olovere 


olover 


klouver 


a,8e. 








grftf 


grove 


grov6 


grouv 


ehart. 








1425 drftf pit 


tdrirf 


drovB 


drouv 


drftfsf 


drovQ 


drove 


drouv 


prftfost 


provost 


provost 


provMt 


a^o. 









ftn 



— onne, a dal Ld. ann si^e 0. oo(n), a(n) Ch, won emphai. And, — 
(w)one, wonnes pi Td. (oon) Q, V, Cp, (wasn) Jn, (<Ni» won 
Dyehs. (wen) Beh, (wvn) Fr. (won) 8K 

nftn tnan none nan 

— ^na(n) Ld. noon : itoon, no Ch, non : Johon il«d. — (00) (?, TT. 

1430 on ftn tan tat anon tenon 

—on an Ld, anoon : onerichoon Ch — (»non, nnien) /». 

eall ftna tall ftne alone eloun 

— ^a<«r alone — (aloon) Q, 

*Anlio onli only ounli 

' ouiqae '— oonll Ck, (oonli, oonloi) Q, (oonli) Jn, 8h, (onli) Beh, 

nftn )>ing tnan ping nothing na]»ix| 

— naming Lay, noting Ch — (o) BU, G — vg (nafin). 

tAne tttnem onoe wans 

Ld. enet (ea) AM, onei Beti, enns Aud, t^ to TJf— (oons) 
G, (wens) Jn, Beh. (o) noi (o), (w«n«) Ld, (worn) 8h. 



1435 ^'^ Bf 


hone 


hone 


houn 


'rook'eAorf. 








■oftn prt 


tehaa 


shone 


Jon 


(•)«. 








stftn 


tstan 


stone 


stoun 


gegftnptc 


tgftn 


gone 


gon, gon 


(00) G, 


{o)Ld. 






gr&nian 


gronien 


groan 


groiin 


1440 drfi^ 


drane 


drone 


droun 


— Prompt, etc. 


»oo. 






bftn 


bon 


bone 


boiin 


h&m 


tham 


home 


houm 


lAxn 


lorn 


loam 


loum 


hwftm dat. 


twhamm 


whom 


hu¥r(m) 



m §W8, VP, Du,, Mm. a ^TTS^wham (was), wluem Lay, hwam AR. 

Z Z 



340 HISTOEY OF ENGLISH SOUNDS. 

hoam Ajf. qaa(i)in North, wham AUP, whom Ch — (whom) 8m. 
(whoom, whaam) 0. (wh«m) Pr. (haum) Cp, Jn. 



1445 fftm 


fom 


foam 


foum 


fto 


ok 


oak 


oiik 


&-oumba 


— 


oakum 


oukem 


Btrftoian 


Btroken 


stroke 


Btroiik 


Bp§aa, 


Bpoke 


■poke 


Bpouk 


1450 orftcettan 




oroak 


kroiik 


t&oen 


ttdkenn 


token 


toukn 


ftlite prt 


oiihte 


ought 


ot 



'poflsened'— ahte, aute (abte) Lay. ouhte (ahie) AM. a^te BOl. 
awcte : bitaucte Hv. aaghte BBC. oghte : bro^hte Ch — ocht SV^. 
(owht) BU. (oouxt) G. (oot) Pr, (oot) Cp, Ld, 

ftgan tajhenn owe ou 

'ponees' — ow«n (ahen) AB. 0500 Ay. awe North., TM. ogen GE. 
owen Ch, 

ftgen tajhenn own oun 

— awen (o) Lay. owen (ahen) AB. o^en Ay. awin : drawin ptc 
(ann : draun) C3f. awen TJIl. auen AllP. owen Karl., Ch — owne, 
awne Td. (ooun) G. 

M55 t^ tlah low lou 
(loou) G. 

dftg douh dough dou 

— do3 Ay, — (doo) dowe Cp. (doo) Ld, 

ftte ote oat(B) out 
(ootB), (wvts) barbare Cp, (otmiil) oatmeal Ld. 

hftt th^t hot hot 

— ^hoot Ch — (whiter) harbare Cp. 

Bm&t prt smot Bmote tsmout 

1460 w&t vb tw&t wot iwot 

— woot CA— thon wottest Td. 

wrftt prt twr&t wrote rout 

g&t ' tg&t goat gout 

b&t bot boat bout 
(booein) Bch, (boosn) 8h boatiwain — (bongn). 

-hftd t-had -hood -hud 

— wreccehed Ld, prestehede : lede vb CM. godhede, maDhod(e) Ay. 
— (-huud) G. (-h«d) Cp. 

1465 r&d prt rod 

i rode 

* riding * — rad North. 
l&d sf tlade lode loud 

' leading/ * path ' aUo in loadstone. 

Bc(e)ftdan tshaddenn ehed . Jed 

■eparate ' — ihadd pte 0. 
Btrftd prt Btrod Btrode etroud 



rftdsf . ^ 

rad 



rode 


roud 


road 


roud 


raid 


reid 



FIRST WORD-LIST. 



341 



w&d wod woad woud 

1470 gftd sf gode goad goud 

tftdige tode toad toud 
— Udde AB. 

bftd abod abode eboud 

* waiting.' 

brftd tbrad broad brod 

(00) Sm, a. (00) Cp, EO. (00) Bck, (00) Sh, 

r&p trap rope roup 

1475 aApe 8ope soap aoup 

sw&pan Bwopen sweep aw^p 

prt swSop. aswopen pie Ru, — AB, Ch, prt sw8p— (ii) Bll. 
grftpian. gropen grope group 
pftpa pope pope poup 
pop HVff. 



A. 

•A tea aea aU 

(Me) (?. (ni) W. (Me) Cp. 

1480 Ar tAr ere teer 

— Br adv 0, ear Ld. nr, ar (are) Lay, fer, ar adv B&l. fare 
North., TM. er Harl. or(e) Aud. er (op) Ck^yet Td, (ear) &. 
{eet) Cp, (iir) EO, Sh. 

rSran reren rear rler 

tskAr akere sheer Jler 

'pure.' OEadr — a2«o 8(c)here. §ii/romOE. 

hAlan fhaelexm heal .h^l 

(ee) Btt. 

thAlo . hel]> health hel]> 

— hele 0, Lay, hele Ch, heele Prompt, hel>e Lay., Ay,, OE, Prompt, 

— (ee) 0, 

1485 d«l tdel deal d^ 

^-del, todeled, daeleth Ld, dslenn vb 0. 
:)vAl ynX thrall t)>rol 

yeSH nom. Jnel OE, from Seand. — trel Lay,, AB, Ay, pi ^rallea 
Lay,, yreUea AB, ^nX North,, f BBC, thxalralCA. 

Alo tillo each HtJ 

7, oe VP, not in Da,, Bu, — elo, e Ld, mlc()i\ elch, ale, alch, nlc, 
ech (ech) Lay. euch /«!. ilch AB, ilk North., OE. ilk, ich TM, 
uch AUP, Hart,, Aud, eeh Ay,, CA— (eetj) 0, (iidj) Ld. 



hmp 


he]> 


heath 


Wj> 


hA]>en 


th»]>exin 


heathen 


hUtSen 


seA> Bf (1) 


tsh»]>e 


sheath 


M 


e, ea,e/Trfi^. 








1490 wrA)» 8f(1) 


wre)>e 


wreath 


r«l> 



342 HISTOET OF ENGLISH SOUNDS. 

wrA))()>)o twra}>)>e wrath fro]) 

(wra» (?. (roo|>) £0, Sh. (i»e» Bch, 

:klA)>dlprt oladde olad tUad 

OE ol&)rian— clojwn and de^n in ME. olAd(e) CJIf. ded : lad pH 3£H. 
cled:bed!rjf. ciad AUP. ydad C%. 

hAs sf biha»0te behest tbihest 

— ^behsm pi 'pramiiM' Chr 1093 — Lay. bihette AR, Ch. btth«fte Jy. 

)4^-lA8-)>e leste lest lest 

to VP. se, e i2u.~leBte Zay. leete (ea) AR, leet j4Z/P, JTaW., TTtc^. 
last, e iiiui. naihelees : peei CA — ^nevertheleMe Td, 

1495 tAsan tosen tease tQa 

*oiupo' — a2ioai. 
lAflsa tlasae less lee 

m VP. e Ru. — ^Ueaie (a) Lay. e, ea AB. lassei-nMse ON. e fjVor^., 
triV. tOi?. -4y. % RGl, AUP, -fAud. +e.t»CA. 

flAso fflssh, fleseh flesh flej 

» FP — flesc Lay. flesch /uZ. fleacha, yleschs, fleohs (flesch, flea) A R. 
fleao, fleia Horn, fleaaoe .K9. . uleaae Ay. fleia, fleaae : leaae CM. 
fleyse : liknea, fleya : neys adj MH. fleis &E. fleiah ^arZ. fleach 
AUP. fleiach Fic/. fleaah (ei) (7A— fleas(h)e 2\2. 

lAst tl»8t least l^st 

leaaeat Du. Isaaeat Ru. — ^le(i)at : be(i)8t North, leeat Ch — (ee) 0. 

lAstan tlasstenn last laast 

'perform' — e, a Ld, ea, e Jul. e Ay. a NortJi. fe G^J?. 

1500 wrAstan wresten wrest trest 
(hreat) Ld. 

wrSstlian wrestlen wrestle resl 

IWS alw wrezlian, wraxlian— e Lay. e Kath., OE. a AR, Ch — 
(wreaU) BU. (real) Jn, Bch, Sh. 

mAst tmast, m most moust 

se VP — ea, e Ld. allmaaat 0. te Lay. ea, e /it/, e AR, R&l, Ay., 
Earl, a North, o GE. fe, fo RBC. fmetmt CA— utmooat Td. 
(00) Cp. (n) Mg. (uu) ^0. (o) jBcA. (00) i5A— (atmoat). 

gAstlio gastll ghastly gaastli 

Grein — (gssli) Jn. 

lAwed tlaewedd lewd l(J)uwd 

* lay (man)* — flawed Ld. lewede (er^) JtU. ]au(e)d North, logede OB. 
lewd TM. leud Prompt, lewed CA— (en) G, Pr. 

I50g tslAw]) 8lou)>e sloth 8lou]> 

alftw 0(2; — alanVe Lay. alouliffe AR. Ble(a)u])e Ay. aleuthe Marl. 
alouthiiKd aloth, dewth, alawthe : trawthe TM. alouthe : trouthe 
CA— elewth Td. 

Afire tfl»fire ever ever 

— oaer Ld. sfer, e(e)uer (eaere) Lay. awre, euer TM. euere CA. 
— (ever) G. 

"'Afire-Alo everich erery evri 

iurio, seareum wile Ld. 8euer8Blc(h), aueralch, euerulc Lay. eauer- 
each Kath. euerich, efrich (euoh) AR. euereich Ay. euerilk 
North., TM. euerilc GE. eueruch Harl. euerich, euery Ch — (everi, 
evrei) G. 



FIRST WOBD-LIST. 343 

lAflm leven leave l^v 

Uving HVg. 

lAfde prt tlefedd pic left left 

— IsBuede Ld. Isfile, », (e) Lay, leafde Jul. leaTed pte AB. left 
North., fSBC. flefte, laft : iluhft TM. lefbe, », ylaft : craft C9k— leeft 

1 510 nAfire tnibfre nerver never 

— nefrft, neure Ld. neauere Kath. never AB, nenre Ay. neuer 
iVbftA. nawxe TM. ner ilaW. 

hlAfiUge tlaffdij lady leidi 

» FP— le(a)fdi /ttZ. lauedi, leuedi JTi^. Iheaediily. leuedi (e, a) CJf. 
lefdiJIfir. leuedi, ledi £ar;. ladi CA. 

Anig taiii5 any eni 

— ani, »ni Ld. »ni, »i, ei Lay. e(a)iii (ei) /u/. oni Prod, eny Jy. 
any North, ani 6^. eny ^aW. ane Aud. ony TTtW. any, feny 
(o) CA— eny Td. ani O. (») jBcA. /SA. 

hlAne lene lean l^n 

— Ihene Ay. 

Iftn sf lone loan lonn 

to lane Sf^. chart, laneaang Wgl, generally ISn. 

1 515 lAnan tlenenn lend lend 

— 8B Lay. ea Kath. lenen CK leendin inf Prompt. 

lAned ptc tlenedd lent lent 

—lent CM, PPl. 

xnAnan tmenenn mean mijn 

' mean *^-aUo mttoen. 

mftnan menenn moan moun 

'complain' — », e Lay. ea Kalh. menen v&, mone tb AB. manen min 
mon Priioner^i Prayer, monen OS. 

mftned ptc mened meant ment 

1520 gemAne timan mean mUn 

' common.' Aet gemftna — ^inuene, o Lay. imeane Jul. 

elAne tclene clean klUn 

dane odv— cbenneae, olennlike 0. m (ea), clane (ea) adv Lay. ea AB. 
e, ie Ay. 

wrftnna wrenne wren ren 

wema Cp. wrenna Wgl — wranne : monne ON. 
clAnaian tclennaenn cleanse Idena 

daanian VP.—e K8, North., Ch. a Aud.—{i\) Bek. (e) Sh. (kliinU) 
deaDly Bdi, ^A— (klenU). 

_ ^^ ^ f emmet temit 

Amette amete < ^ 

\ ant aant 

— BCH. amote Ay. am(p)te Wid. aUo emote, emete, ematte, eramotte, 
ante^(»nt) ant, (sent) aunt Ld. (lent) Beh, Sh. 

1525 Amet(t)ig empti empty emti 

* unoccupied '—empti AB. emti Ay. amti BOl-~(empii) G. 
glAm glem gleam gUJm 

rAcan reohen reach r^tj 

— -recfae ^oitA., fTM. 



344 HISTORY OF ENGLISH SOUNDS. 



hrAcan 



r reach rytj 

( retoh ret/ 



tSoan ttochenn teach tytj 

* show * — teche, he tek^ Ay, 

1530 blAcan bleohen bleach biytj 

tAhte prt ttahhte taufirht tot 

— eA, a, e Ld. teebte, itaiht, tahte, taute (tehte) Lay. tahte Jul. 
teihte AB. tachte KS. to^te Ay. taght North, a Q£, Harl. 
taughte CA— (Uuht) 8m. (tooxt) Q. 

BtSger ateire atair ateer 

wAge weie wey wei 

'weight/ 'acales.' 

hnfigan nejen neigh net 
(nsei) Pr. (nii) Bch. (nee, nnn) Ld, (nee) 8h. 

1535 cSgfif keie key kU 

— keie Ld. keye : pleye Ch — kae HVg. kayei Td. (kee) Pr, Jn. 
(kii) EO, Ld. 

cIAg olei clay klei 

— Ch — (ai, aai) Q. 

Ag-hw»])er te^sJ^err either aitSer,ytSer 

aUo ffg^. 6g^ Pu. — niVer, ei (ai) 2^. eiOer AH. eider, ai ily. 
ayther North. a(y)ther TM. eyjwr Harl.—(fi\) 8m. (eei, ei) G. 
(ei, ee)/n. (e) Ld. {ii) EO, 6h. {n) Beh, Fr. 

*ii£ghw8B]ier tnoww)>err neither naitSer 

nother Ck. (eei. ei) G^. (ee, e) Cp. (oi, ee) /•. (ee) EO, (e) W. 

(ei) Bch. (ii) <S% — «ee nft hw»]7er. 

h&to th»te heat h^t 

1540 ^afiti sn tuBte aeat ajjt 
(oe) W. 

awStan aweten aweat awet 

swftt eft— Bwit 8b 0. zuot «& Jy.— (ee) 8m. (e) BU. (ee) Cp. (set) 
Jn. (swot) r^. 

hw&te twh»te wheat whyt 

apAtte prt apatte apat apaet 

spetan inf. — speten tf^. 

fStt tfiEktt fiat Itot • 

»FP. 
1545 lAdan tledenn lead lj|jd 

— e L(J. 86, ea. e (eo) Lay. ea Jul. — leding H Vg. (ee) Wk, Pr, 

hlfider ef laddre ladder leder 

leddre AB, OE. Ihoddre Ay. a RGl. 

aprftdan spreden apread spred 
eeCk. (e) Q. 

lAdde prt tledde led led 

— e Ld. e, ea (a) Lay. ea (e) AB. e fNorth., \Hv, GE. led : bed. 
bid : had i2£C. a Harl., Aud. ladde : hadde CA— leed Td. 



FIRST WORD-LIST. 345 

sprAdde prt tipredd ptc spread spred 

— a Lay. ea (e) AB» e -fNorih., "fSBC. a Aud, spradde : hadde Ch 
— spreed Td» 

1550 gemAdd ptc mad mad mad 

gemasdid Cp — ^madd : radd ('^01 hrsdd) CM, medde : ledde prt HH. 
maAigUdRBC. 

baddel badde bad bad 

' hermaphrodite.* 



& (e). 

gA (5) tga yea ijei 

goa TF)9. gee,g& Du. go Ru, — ^eaiGsM. z^ ^ni AB. gia(yaa):Bua 
CM, ya Prompt, — (ee) 8b. (jee, jii, ii) Jn, (joo) nutte Cp, (jii) 
EO, Ld, (jee) Bch, 8h, 

Arende (d) termde errand erend 

— SBrnde, er(e)nde, amnde (ea) Lay, erand North, erand, arand TM. 
erd(e)ne OJE, arende, ernde AllP, 

bAr (d) thar hair beer 

her Du„ Bu. — ^he(a)r Eaih, fhare North, hor : lor, her : >er Hv, fhore, 
thare TM. here AllP, heer Wicl,, iCh—ee Td. (heer) BU, (ee) 
Cp, 

1555 )^r (5) t]>ar(e) there Seer 

e VP, Du, » Bu. », a IWS — ]wer, ea, e, a, >iBre etc Ld, Lay. 
>e(a)r Kath. >er(e) AB, Ay. t>ere BGl. )>are : fare i;& ON. ]Mtr(e) 
^or/A. tCer, Cor OR fe, +0, fa BBC, TM, t>ere -4«P. Jwre: 
more, a, e Earl, thore : lore sh, e J ud, — (ee, aa) Sm. 

wAron (5) prt twarenn were weer 

e VPf Du,, Bu, rarely » in Du., Bu. — wieron, we(a)reii, a Ld, 
we(o)ren, a Lay, weren AB. e, a K8, e Ay, fwere, twar(e) 
North. te,+a^4/?P. fo. e G Jf . to, fe TJf . t», +e i2-BC— (weer) 
(7, Cp. Jn, i?0. (e) Beh, 6h, 

bwAr (5) thwaar where wheer 

e VP, Du, m Bu, », a IWS—Bowwhtir 0, hw(e)ar, hwere Ld. 
whser, e, iwere (ware) Lay. a, (e) ^122. e Ay, qufir(e) North, quor, e 
OE, quere AUP, t»i to, « TM. t». t* iJ-BC—hwier HVg. 
(wheer) G, 

fAr (5) titer fear flor 

'danger* — offiered Ld, offearen, fe(o)rlich Jul, — feare tb, fearful Td. 
(feer) G. (fiir) Cp. 

gAr (5) tsar, e year Jier 

gear W8, e i)tf.— gear, gser Ld. jer Xay. ^eare c2a<. Proc/. gier 
(yeire) CJJf— <!i) -8^. ^^, Bt. (jeer) G. (jiir) Pr. 
1 560 bAr (d) Bf tbare bier bier 

eeDtt. ejRtf. {i'l) Sb, 

Al(d) el . eel Ul 

ge8Alig(5) tselij tilly siU 

* beatas '— -eele Aud, lely (ee) Ch. 
mAl (e) tmal meal mJJl 

8b implieB (miil). 



346 HISTORY OF ENGLISH SOUNDS. 

brS)>(*e) bTe)» breath hrep 

~bre» (ea) AR. breeth (e) : heeth C%— (e) BU. 

1565 *brfi]>an bre]>6n breathe br^tS 

(ee) G. 

"'hwAsan (*d) wheaen wheeae whQa 

hweofl prt JBfeH» 

oAse (5) ohese oheeae tj^a 

^lW8—m,^JA, tsiB /Si. 

blAst (5) blast blast blaast 

—SB I«ay. 6 Jy. a Bt^., fGE, fTM—{eBBB) Cp, (se) Beh, Sh. 

mAw (e) xneaw 8ea(inew) txpi^li^ 

6 Ep., ea Cp. » TTiS^mowe Proinpt. 
1570 Afen (e) fefenn eTen(i]ig) Uvnil) 

efem Du,, efen JZu. nfenung jB/oH — euen Lay, — (iiTnig) G. 
lAoe (s) tlAche leeoh lljt/ 

— « J^ — (ii, ee) Sm. 

aprfic (d) fspsohe speech ipUt/ 

spnc IWS, spreo Du., J2tt. — spneoe cM earn Ld, ipeche (e) AJl. 
speke, Bpeche : meke adj ffv — (ii) Bll. 

twAg (e) wawe wave weiv 

— ^wage (Xay.). ^y. wawe AUP, Wid., Ch, wawghe TM, infi of 
wagian {ME wawien) — tt|/f of wafian. wave Td, waw Sb, 

hwAg (*5) whei whey whei 

iS75grAg(g) grel gray, grey grel 

graye, grey Td, (eei) Pg, (ee) Pr, Ld, 

At prt (5) t6t ate eit, et 
— eet : feet Ch, 

lAtaii(d) tUbtenn, « let let 

— latenn 'behave' 0; 01 l&ta. m,^ Ld: e Lay, e (eo) iliZ. a (a, e) 
CM. leete : strete MH, loten : bihoten GE. o MP. leete : heete 
( e^hvte), imper, leet. lat (e) CA— let(t), lat Td, (r) harbart Cp, 

strAt (e) sf tstraete street strQt 

— Stretford Ld (stnatfed, stnefed) Stratford dc, Strafford. 

wAt (5) weet wet wet 

— « Lay,, AR, wate PC, woytt TM. 

1580 mAte (d) tm6te(llke) meet tmtJt 

* moderate ' ^met H Vg, (ii) G, 

blAtan (d) tbl»tenn bleat bmt 

rAd(e) trad rede trQd 

'advice* — s, e Ld. alio ra]> 0; 01 rS>. « (ea) Lay, e (ea) iUZ. 
rathe vb : ba>e * both* Hv, frede TJtf. 

rAdan (d) tradenn, 6 read r^d 

e RU. w Ru.-^n Ld. redd ptc 0.— (ee) non (ii) (?. (ee) W. (ii) 
Cp. 

rAdelB(*d) redels riddle ridl 

—also i— redlee pi Td, 1 

i585]>rAd(e) }>red thread pred 1 

— ee Lay. — (ee) 8m, 



FIRST WORD-LIST. 347 

■«d(«) tsed seed s^d 

iiAdre(6) tneddre adder »der 
iuBd(d)re ZTPS^-nadre Ld, neddre AR, BeH., GS, neder (dd) CM. 
«ddr9 Ay., Wid. edder TM, naddre Ch, 

xiAdl(d) tnedle needle ii:Udl 
nmiSl Ep,9 netbl Cp — ^nelde AR, 

^ (medwe meadow medou 

'meadow.' een, mmdwe—mo^eiwe dat Lay. medewe Prompt, mede 
Ch. 

1590 grAdiff(d) tgr«di$ greedy gr^di 

(ii) a. 

d«d(e)8f tdede deed d^d 
>-4iedbote 0. », e Z;<i. dede iLS— did HVg. 

on-drAdan (d) tdradexm, e dread dred 
e Du., Ru. — t, (e) AR. pU dredd 0. pH fdradde, fa Ch — (ee) Sm. 

bl«dre(d) bladdre bladder binder 
bbed(d)r6 IWS^hleddn AR. a Ch. 

rAdde prt (e) tredd ptc read red 

— geredd Ld. a Lay. e North, a j^Harl., Ch — reed Td. (e) 
Sm, O. 

1595 elApan (d) talapenn eleep sUJp 

e, a W8 — abep #5, ilepen Ld, abep, a, e «b 0. abBpen, eo, e, (e) 
Lay. 

8oAp(5) tshep sheep J^p 

ea WS. i Du. e, lo Ru. — icheap, e (^) Lay. acbep AR. nep Ay. 
■oepe : kepe 96 CM — (ii) 8b, 6m ete. 

soAiHhirde (d) sheepherde shepherd Jeped 
aoheephcord Ck. (Jepherd) O. 

wApen(d) twspenn weapon wepen 

— ^wapen, we(a)pDeD pi (wepne) Lay. wopen, e OE. 01 Tipn — 
(wiipn) EO, Ld, Bch. (wqpii) Sh. 

sIApte prt (d) tsleppte slept slept 



1600 M the he 14J 

— (h)e (heo) Lay. heo Horn, he, lia KS. he, ha, hi Jy. — (hii) G, 

)« t)>e thee tSU 

w6 twe we wU 

wi JF^. (wii) Pr, Cp etc. 

vA tme me my 
(mii) -Pr. (niM) Cp, Jfjjr, Jj». 

gd tje ye UU 

—ye, hye KS, 






348 HISTORY OF ENGLISH SOUNDS. 

1605 Mr th6r(e) here hier 

— ^herrbiforenn 0. her(e) Lay, her(en) Horn, fhere RGl, MBC. hier, 
hyer Ay. — (hiir, heer) Bll. (hiier) 0. 

geheran fherenn hear hier 

ie, i IW8, J lW8—henJL (u) Lay. u BGL e, (u) PPL here, hyere, 
hiere^V. haratow (- berest \iVi) TM—heue Td. (heer, hiir) BL 
(hiir) Wete. 

etSran tsterenn steer stier 

7IWS. 

woBrig^ weri weary wieri 

win ffVg. (weeri, ii) Bt, (weeri) O. (weri) Pr. (wiiri), (w«ri) 

harhare Cp, 

brer(e) tbreress pi briar braier 

brere : manere, breree : geres ( ' dreai ') Ch. brere : chere TM — ^bryres Td<. 

16 10 blere bler(eyed) blear(eyed) bUeraid 

blerie«blerige < blear^yed ' (?) cAar^.— (bliir) Pr etc. 

,. ^ ftherronenn hearken thaakn 

hdronian < 

\ herkien hark haak 

— herkyn, barken ; barke imper. TM. herkien Horn, berk imper. : werk 
BBC. herkin Prompt, herken CA— (beerkn, a) Bt. (barkn) G. 
(Iierkn) Mg. (berkn) EO. (hnerkn) Bch, 6h. 

gehdrde prt the(o)rrde heard heed 

y lW8—heirde, a, i (o) Lay. u BGl. e (u) PPL yh(y)erd Ay. e, a 
IfoHK. fa TJtf^berde, a Td. (aa) G. (ee, a) Bt. (») Pr etc. 
(e) Jn. (ii) Ld — (hiad) vg. 



hfila 


hele 


heel 


hi]l 


Btdle 


Btel 


■teel 


etui 


ylW8. 








fdelan 


felen 


feel 


fljl 


~ - (ii) G. 








tfSlagi 


felawe 


fellow 


fidloi 



I6I5 



— feolahe (feolobschipe) Jul. fe(o)lawe AE. vela^e Ay, felaghe North. 
felage GE. felawes : dawes RBC. felowe TM, WieL felawe : shawe 
CA— (feloou) G, 

tee}) pi te]> teeth ti|j]> 

brGl)>er dat. tbre}>re pi brethren tbretSrin 

pi br6)H>r, gebrO]>ru — ^bre'Sre pi Ld. bro]ieren, brejwren (broJierB) Lay. 
hrefSrenAB. hte^ North. bieOere GJ&. brebher Tif. britheren, 
e WieL breetherede Ch — (br^ren) aut (bretSem) G, 

goM pi gee geese gUs 

gyaie Td. 

1620 brdsan brusen bruise bruws 

J ZTFS— brimn, o Prompt, u (o) Ch. brened pte TJf— broeed Td. 
broosed Ck. (in) Jn. (uu) Z4. 

wdeste twesste waste weist 

' deeert * — weste Lay. wast : mSst CM — (aa) 0, 

gerd&fa reve reeve r^JT 

— grej^fe 0. from Seand. greifi. 



FIRST WORD-LIST. 349 

Bcir-gerofb sohlrreve sheriff Jerif 

— Mhiireve : ilSve Moral Ode — (▼) Ld, 
geldfEui tlefenn believe biiyv 

bely&n J^/eJET— bileven Lay. bUefde Jul.^(i\) G, 

1625 slfif sieve sleeve slijv 

a/«o defe (f). jlWS. 

p^tp Bf "pefpe theft ]>eft 

y lWS--eo Lay., AM, u £Gl. ie Ay. iS^te OS. thefte CA. 

hdng prt theng hting hax) 

— « Lay. o North, e, ee (y, o) Ch. 

gesdne adj tsene ) 

gesewen ptc tsejhexm j 

gasSne wed as pie in Du. and Bu.; gemene, ea aUo Bu. j IWS. 
geaewen WS. gesagen VP — sen pte North., Betty AUP. seyne, eayne : 
oyne * eyes ' Aud. leyn Wioh yaeyn : ayeyn, sayn : fayn uA — (i) Jn. 

sodne fshene sheen tjijxi 

(i)e, y, eo, Uy—€deo 80(Sne 0. ■c(e)one (e) Lay. 

1630 woenan twenenn ween twQn 
(u) iSb. 

ocene tkene keen k^n 

owden sf fowen queen kw^n 

—also cwene— kwin HVff. (ii) G, Beh, Sh. 

gr&ne tgrene green grfjn 

(ii)Gr. 

t6n tt^ne, tenn ten ten 

y IWS— ien(n) Ld. ien{e) Lay., AB. Uai Ay. tgen : men GE. ten: 
men TM—{i) Mg. 

1635 -tSne t-tene -teen -tyn 

J IWS Oirtin, |rirtiin) G. 

gesceman tsemenn seem sUm 

' reconcile ; * 01 Bftma * befit ' — (biaiim) G. 
tfiman ttsmenn teem tQm 

y IWS — injl of t6am sbsi. teamen HaliM. temen ' prepare ' Lay. 
doeman tdemenn deem dUm 

(U) G. 

^brcemel brembel bramble brsmbl 

e, y — ^brimbil, bremmil Prompt, brembil, i Wiel. — (a) G. 
1640 Scan tekenn eke Uk 

ilWS (iik)(7. (eek)/«. 

rOo rek reek r^k 

SGBcan tsekenn seek sQk 

— secben, r Lay. eechen AB. seche, he zek]> Ay. sek North, oh 
Harl., AUP, Aud, k (ch) PPl. k Wtd. fieke, fseche Ch. 

be-8GBoan beaeken beseech bisytj 

— ch Lay,, AR, KS, Ay. be8eke>, biseche)> KS. k North., fBBC. ch 
Barl.,AUP,iAud. +ch, fk CA. 

beohe beech b^tj 



350 HISTORY OP ENGLISH SOUNDS. 

1645 bdcnan tbeonenn beok(ozi) bek(en) 

i IWS—htdCDien Lay. bekke : Senekke Ch. bekiiin, bekin 'nutare* 
Prompt. 

- 1 Ijreoh (breech tbrOtf 

^ ( breeches britjls 

— breke T3f. brech^ brek 'braccae' Prompt.— (bziitf) 8m. (britfes) 0. 
(britO Ld. 

heh]>o sf heighte height halt 

i lW8^he^]>e Ay. heyt MH. heighie : feyghie vb RBC. hyght : mvght 
TM. heitbe Prompt beigbte (e) C^— haight Ck. (heixt) 0. (heit, 
h«et) Cp. (bait, beet, beet» Jn, the last spdt heigbth. (bai]>t) Ld» 
(beet) EO, Bch. (hoit) 8h. 

ndhst tneest next nekst 

y IW8. n«Bt VP, Vu. nabst, i Bu. — ^nextd Lay, nixte sh Ay. natfalde 
Jul. nest : prtet ^or<A., GJS'. nest fRBC, Harl. 

*Sga]> eit eyot eit 

WS igga]>, igeo]) — eit Lay. eit /rom *SbJ>, ♦SgJ> {ep eiht /rom ge8ib]»). 

1650 6g-land Hand island aflend 

ilW8. ealond FP— eitlond (ilond) Lay. eilond Bet/. 

hSg hei hay hei 

ilWR (bei)5Zl. (bid)Gf. (bee) Q). 

legetu sf leit lightning laitnii| 

i IWS — Lay.f AR. ai Ay. leiten vh, confused w. libbtenn 'illuminate/ 
libtnen * abine.' ligbtninge ' fulgur ' Prompt. — (leixtnii)) 0, 

tsloBg tsleh sly slai 

— Bleytney 'near' RGl, sle^e pi Ay. 8legbe:degbe («"d6yja) CM., 
ily MB. slee^i sli^ Wid. sly : hertely Ch. 

wrcBgan twre5(h)enn (be)wray tbirei 

— ^wreien AB, Ch. wraie Ay. also wrie. 

1655 tegan teien tie tai 

i IW8 — ^teien (ti^e) Lay. tden AR. 

twdegen ttwe^^enn twain ttwein 

dSgan deyen dye dai 

i / WS-'Ch. dyyn Prompt. 

Jsloeg]) sf sleihte sleight tslait 

< cunning* — liste (sleb^) Itay. sle^Jie Ay. sleigbt : beygbt (ab6b^) 
BBC. aleigbte : eigbte, slygbte : mygbte C^^ai) Ld. (sLoit) Bch. 
(slait) 8h. 

Idtprt tl6t let let 

— le(o)t Ld. le(a)tte Lay. lette AB. leet North., GB. leet : feet Ch. 

swQt 

fljt 

mjljt 
gpyt 



1660 swoete 


t8w6t(e) 


sweet 


scdte 


shete 


sheet 


jlWS. 






foet pi 


tf(§t 


feet 


-~(H)a 








meten 


meet 


groetan 


tgretenn 


greet 



FIRST WORD-LIST. 35 1 



1665 


Mtel betel 

Mualleui/ jlWti. 


beetle 


bUU 




gemoBtte prt 

*liOBdan 
hadan. 


metta 
heden 


met 
heed 


met 
hjyd 




stoda 


Btede 


steed 


tstil]d 




speed sf 
•wealth.' 


taped 


speed 


■P«d 


1670 


foBdan t&denn fted 

ffiding EVff, 

ndd 9f tnede need 

16 eW8, y, 6% eo lW8—Jkdod(e) Lay, nede, 

mM if tmede meed 


fUd 

njjd 
neod AB. nyede Ay. — 

imUd 




crida 


erede 


oreed 


krtid 


1675 


^brodan 

e (ii) 0. 

*biadaii 


breden 
bleden 


breed 
bleed 


brtid 
blUd 




spcedde prt 
fcBdde prt 


spedde 
tfedde 


sped 
fbd 


tsped 
fed 




bloBdde prt 


bledde 


bled 


bled 


■ 


blcsdaian 


tblettaenn 


bless 


bles 



blediUui VP. bl(o)edBla Du, bl(o)etnAn Bu, bledsian, bletsian W8. 

SBblltMwl eire— bl«tond Ld. aUo blettoedd 0. bletieigen Lay. 
leMden AB. hhmed Ay. hLvmed, blifloid North, fblyst pie TM. 
bliaoed QE, eblestibest Avd. bliiie:kiMey bloMe : ounednene Ch. 
ff^ of bliM— blMijnge, 7 Td, 

1680 stSpel stepel steeple stUpl 

jlW8. 

*strdpan strepen strip strip 

7 — a Jul, e Ay, e (ee, i) Ch — itripped Td, 

WGBpaa twepenn weep wtjp 

*CGBpan tkepenn keep kUp 

6. 

tdflpe depe depth dep]> 

7 lW8-4Jh, depthe Wid, 
1685 *<K»Pta tkeppte kept kept 

e. 

0a. 

flfia(h)8f He flea flU 

— (u)Tr. 

pte peook pea(oook) pUkok 

Grtin, generally pawa— o Ay. e (o) C/»— (poukok) Pocock. 

darsn er(e) ear ier 

* ipioa.' eher Dm. lehher Bu.— «aMt pi AB, QE. 7«re Ay. 



352 HISTORY OF ENGLISH SOUNDS. 

6are taere ear ier 

— ^yere Ay. — (eer) 0, (eer, iir) Bt (iirwig) Cp. 

1690 searian seren sear tsier 

(ii) Cp. 

ndar cp tner near nier 

— ^neor Lay^ Jul, nere : lere t>b CM. nere : here adv, nar : war a4i TM. 
neer (nere) Ch — (nilr) 8m. (neer, nier) Q. (niir) W etc. 

geara jore yore ijor 

iara Ru. — ^3(e)are Lay., AE. yoore Ck, 
t§ar tt»r tear tier 

tear VP. tear, tsher Du. (ii) Cp. 

leaSor — lather l»tSer 

leaOorwyrt Wgl — ^le]76rede a swote Lay. Ii0erede KtUh. 
1695 dea]> td8B)>(}>) death de]> 

— ded North. de(a)d : forbead GE. fded Hv., BBC. fdede^ fdeihe 
TJtf— (ee) 0. 

-leas t-l»8, -less -less -lis 
leas (e) AB. -lees Ch lease Td. 

oeas prt tchies chose tjoiu 

• — cbses Lay. cheas, ohesetese $b North, fcbes G£t BBC. ches, 
choB : porpose AllP. chos Harl. chase, fe TM. chees : doutelees 
CA— (00) G. 

east taest east iljst 

— test Ld. yeast Jy. — (eest/eep) J^tcheap Jn. 

£astron pi tnstre easter iljster 

— esterne dat. estren dat. Ld. 8est(e)re (easter) nom. Lay. iestre Ay. 
esterne GB. eestem Prompt, astere And. 

1700 h§awan th»wenzi hew tl;|uw 
(heu) BU, Bt. 

hreaw . rau raw ro 

]>§awa8 pi t]>»wesB thews tj^uwa 

' morals/ 

soeawian tshffiwenn show, shew Jou 

* survey * — soawe Ld. soewede, e, a (sewede) Lay. schea(u)weny 
Bchawen (schawi) AB. seaweth, seywin^e KS. 88e(a)wy Ay. sea we 
(scau), scawid (sceud), scaunpto : draun CM. shewe : Bertnelmewe PC. 
schau : knau, schewes : thewea ( s-J^gawas) MH. shew : thew, show, a : 
knaw TM. shauweiknawe Hv. shewe :rewe 'row* Harl. schew: 
know Atid. shewe :fewe Ch — shio (showe) BVff. shewe Td. (eu, 
Jbouz) G. (uu, eu) Cp. (ooa, ou, iu) Jn. (iu, 00) Ld. (00) Bch, Sh. 

screawa shrewe shrew Jruw 
(Jriu) £0, Bch. (Jruu) Sh. 

1705 streaw strau straw stro 

streaw-, streow-berge Leechd. strewu pi Wgl. streu Bu. — bedstrau SB. 
strea AB. stra : ga vb, wa Hv. stre : wee ( »> w6a) Harl. stree : thre^ 
straw Ch. strau PPl. strauberi Prompt. — (au) Sm. (oou) G. 

fga(we) tfiewe few ijuw 

— §BBa, feuna dat. Ld. feue Lay. feawe Kath. ueawe Ay. vawe, fowe 
BGl. fone North. : by anal, of hwOn. fo : wo GE. fo : to go BBC — 
feawe Td. (feu. feeu) G. (feu, fiu) W. (foo) harbare Cp. 



FIRST WORD-LIST. 353 

dtew fdmw dew 4Juw 
(den) Sm, (deui) dewy 6. 

hdalbd tluafSBdd head hed 

— luBued, hefed Ld, heaued, h»ued, he(f)ued, haf(u)ed, hefd, hefd 
Lay. heaued AR, Ay, heaed : weued (-igewsfed), hefd MH, heued 
GE. hede BBC, TM. heed WicL heed, heddes pi (heuedee, heedee) 
CA— he(e)ddee Td. (e) 8m, (ee) Q, 

fbran-hdafod fbrheved forhead fbrid 
forheddes Td, 

1 710 be-rtaflaxL bireren bereave tbir^Jv 

(ee)^. 

berSafbd ptc tbirafedd bereft tbireft 

— ineved Ziay. ireaved iiy. teh North, biraft : shaft CA. 

Idafen lef leaf Ujf 

— ea, ia Ay. 

Idafsf tlefb leave V^v 

—•from iilfifaii. leave, » (e) Lay. — (ee) O, 

geldafk tU»& belief biUJf 

— ^Qaefe Lay. bileaue AB, bileve Ck — \ibfrom gelSiiin. 
1715 Bodaf tshaf sheaf J^f 

dfiaf tdaf deaf def 

— deef CA--(ee) Sm. (ii) EO. (e) Beh, 8h, 

bdan sf bene bean bfjn 

(ee)(?. 



adam 


aem 




seam 


sUm 


steam 


stem 




steam 


stiUm 


(ii) Jn. 










1720 atrdam 


tttrttm 




stream 


str^m 


tSam 


-Ham 




team 


tym 


(u) Jn. 










drdam 


tdr«m 




dream 


drUm 


'melody/* joy '- 


-OJdraom' 


dream. 


» 

• 




bSam 


tbann 




beam 


bUm 


(ee)G. 










4ao 


tec 




eke 


•HJk 



ee VP, ec, sk: Da., J2if.— «ac (eke) Lay. eke Jul. feek, +eke CA— 
(iik) Q, (eek) Jn. 

1725 hrdao rek riok rik 

ifrom Angl. hrCc. 

ISao lek leek UJk 
(ii) Bll. 

g&r-ldao garleek garlidk gaalik 

c6aoe cheke cheek tJUk 
ea, eo IW8. e /Hi., £«.— cheoke AR, cheake Ay, ee Ch — (ii) Pg, 

bdacen an bekne beacon b^en 
becen VP. beoon Pu. becan £u.— bocknen r6 H0r{.~(bek0iisfijld, ij) 
fieacoiufield. 

A & 



^^54 HISTORY OF ENGLISH SOUNDS. 

1730 hSah theh high hai 

heh VP, Du., Ru.^hBdge pi, hehlioe Ld. he^fae pi O. he(i)b, h«h, 
haih, hah, pi hehe, heia, ha^e etc (hej, he^e) Lay, heih (heh) AR. 
hey : ieey (» geaeah) ROL hejliche Ay. hegh, hei North, heg CrJE*. 
hey : aley wfj Hv, hey : wotpy, ]>ey fley ' flew * RBC. hee : see vh, 
hy : body Tlf . heh : ndi, he^e : dreyje ( » dr^ogan) Harl, hyj AIIP, 
Aud. hi^ Wiol, heigh, hy (high), pl hye : Lambardie Ch — hye Td. 
(haiz, hei), (heier, heiar) ep 0, 

h§ahfore heifire heifbr hefbr 

aho heahfra — aleo haif(a)re. hekfere Prompt. — (ee, e) Jn. 

l§(th — lea tllj 

a^leag. 

)>Sah tjwhh though tSoa 

8B VP, e, ft Du, », ea Ru. ea, e Pa$i. — )»eah, 8b, Jwhunethere, )k>^ 
Ld. Bwajwhh 0. beah, se, e(i), a(i) Lay, >auh (a) AR. ^y £6i|. 
>a5 ily. tho(g)h, thof North, thoug(he), ihof TM. 9og C^JF. ))ah 
JOTarZ. ^laj ii2/P. tba^ iiucT. thou^ Wid, Uio(Q)gh Ch. OT >5 
/rom ♦jKJh, o.— (Coo, Coou) 5m. (Goouh, ISowb) S/2. (tJox)fl'. (9oo) Ld. 

nfiah tneh nigh tnal 

neh VP, Du,, Eu. ne(a)h Ld, ne(i)h, » Lay, neih (neh, nea) AR, 
nei : sley (u(; RGl, nie^ ^y. neix : dxeigh ( i- dr6og-) (nei) CM, 
ney : by adv RBC, neg GS--nB\ EVg, (neiz) O. 

1735 xi§ah-geb1ir neighebour neighbour neiber 

— neihebur AR, iie3(3)ebour. ne^ibor Ay, neighebor (neighbur) Ch — 
(neibor, neebor) Pr, (ee, sse) Ld, (neeber) EO, Bch, 6h, 

§age tejhe eye al 

e VP, Du, e(a) Ru, — eegon pl Ld, segen, ejen pl Lay, eie (ehe) AR, 
e^e Jy. eye:deye v& MM, eyne p/:pyne ^^C. ee:bee fi6 TM, 
e^e i^arZ. yje : dvje T = deyja), ygen pl : i wene vb AUP, ige TFtc/. 
eye (iye, eyghe) : meloaye Ch — eye Td, (ei) Sm, G. 

l§ag 8f leie lye lai 

— l(e)ie Prompt, 

flSag prt tflah flew fluw 

fleg VP—e Lay. ))ey fley : on hey RBC, fly, fleigh, flaugh (fley, fleegh) 
C*-(yy) ^\ 

]>reatian ]>reten threaten) }nret(n) 

(Jreto) B«. ()>reet, ]>reetn) (?. 

1 740 grSat tgr4t great greit 

grett Td, (greet) 'magnus,' (greeet) 'ingens* G. {ee) Cp. (ii) 

£0. (ee) Bch, 8h. 

bdatan tbaotenn beat b^t 

(ee) (jr. (beetn) p(e Mg. 

read reed red red 

(e) 8m, 

Idad leed lead led 

(ee) 8m, 

8cr§adian sehreden shred /red 

1745 dead td»d dead ded 

—dyad Ay, deed : breed (» Sa) Ch. didle Aud, — (ee) G. 
bread tbrad bread bred 

— bread, bryad Ay. — (ee) Sm, G. 



FIRST WORD-LIST. 355 

htap thfl»p 

— ea, ya Ay. — (ee) G. 

hldapan tl»penn 

— ^Iheape Ay. — (ec) O. 

Btdap tatttp 

- - (ii) G. 
1750 o§ap 8n ohep 

' purobase ' — guodcheap ' cheap ' Ay. 

*c6ai>-lkni sf ohaiEEure 

faro ' journey '—oheffare AB. cfaapfare Ay. chafiare : ware ' wares * Ch, 
oSajKxnAiin tohappmann ohapman t|fl»pm0n 

— chepmoDi a Lay. ohepmon AR, chapman PPl, Ch. 



heap 


h«p 


leap 


liJP 


steep 


8typ 


cheap 


tfijp 


chaffer 


tt/aefer 



60. 

hldo le lee ly 

'shdUr/ hleow WS. 

}>r9o "^^^[0) three ]>rU 

eo, ie, 7 W8^tTe{o) Ld, Lay. ]ireo AR. >rie KS. yti Ay. thre 
yorth. — (ii) G. (|)rcpenB) Jn; (>ripiDB) JIfy, B^; ()vipena) 8h 
threepence, (^pini) Bch; (>ripem) 8k threepenny — (^pens); vg 
Orepons). ()vipeni). 

1755 Bdo fern ache she JiJ 

heo ' she * 8S0 * that one/ ' the * — aca Ld. heo^ hoe, he, ha (300) Lay. 
h6o, >6o AR, hiy si Ay, sho 0, flc(h)o:d9 North, ahotdo //v. 
ache : to be, icheo RBC. age, she : ^, g(b)e GE. ache And, »hee : 
beantee Ch. 

gesSon tse(o)n see sU 

— ^he 80(0)1$ Ld, iaeon AR. a(y)e)» pi, yai in/ Ay, ■e(he) : thre Mff 
— ^wi sin BVg, (sii), (aiin) pic O. 

frfio "ffre free frU 

freo, fing TT^— freo 2xf, Lay., Ji^— (ii) 0. 

fldo vb tfle(o)n flee flU 

— fle(o)n Lay. fleon ii A. yle^ pi, beuly in/ iiy. fle : me MH, fle : be 
vhRBC. flee: thee CA. 

cndo tones, cnewwess pi knee nQ 

oneow ir<S. cnew 2>«. cnea i2u. — aUo o cnewwe 0, cneo(u)wen, 
oneon (cnowes, cnouwea) p/ Lay, knee North, knowee, knees Ch — 
(nhii) Cp, Ld. 

1760 gldo gle glee gUJ 

gliig Pati, — gleo Lay., AR, gleu : breu ( « brSowan) MH. gle : nanito 
CM, gleu : greu prt Hv. glow : knew pri GE, 

trdo ttre(o) tree trU 

tre(o)w, treo, gtn, treB,jd trew, treo VP. tre(o), tren Du, treow Bu., 
W8 — treowwess pi 0. treo, pi treo(we)n (troaes) Lay. tran Ay, 
tre North, tre : be, j>2 treen GE, 

MoBf be bee bij 

bdoTb tbe(o)n be b^ 

— ^ben Ld. been, pi beoiS, baO iioy. boon AR, bi(eo) tii/', ib(y)e 

A a 2 



356 HISTORY OP ENGLISH SOUNDS. 

pie K8. by it\f, bye)» pi Ay. buen, be(o)ii Harl, ge bun And, — 
(bi) O. 

hldor sn ler(e) leer lier 

* cheek * — leor AR. lure ffctrl, 
1765 dSor tde(o)p deer dlor 

'wild beoBt'— der, sb Ld, dor Horn. de(o)p Lay. deor, daer Harl. 
— (diier) G. 

dSore adj tde(o)re dear dier 

ie, y TT^— dsere Ld. eo, u Lay. eo AR. diere, dyere Jy. — deare 

Td. (ii) iSm. (ee) not (ii) -B*. (diier, dier, deer) G. (diir) JT «<c. 
(diir, der) Jn. 

deorling tderrlinng darling daalii| 

e Du. y IWS—eo Lay., AR. e Wiel. a Attd, e, a rjtf— derlinge 
Td. (ee) no^ (a) dearling Bi. 

drSorig tdrerij dreary drier! 

— u, (e) Lay. eo AR. dririhed GE. 
beor ber beer bier 

(bier) G. 

1770 f3or]>a tfe(o)rJ>e fburth fbj> 

— veorSe AR. uer))e Ay. e Norih,, GE. fur)»e ITor^ fourt TM. 
fertbe (ee, ou) CA— (oou) G. (au) Pr. (uu) Cjp, Jn, EO. (00) W, 
^cA, Sh. 

fgor]>lizig ferthing fiarthiiig &a8ii| 

feorSung Dtf.— farthyng TM—a Td. e Ch. (fierdiii^ Cp. (fieiBPdin) 
Bch. (fa»iiSiij) 8h — (faadin) vg. 

jistjdm sf Bteme stem Bteen 

' steering/ 

hreol rel reel r^l 

hweol 'hnrhe(o)l wheel wl^jl 

also hweowol, hwe(o)gol — hweol AR, huegel Ay. 
1775 geola +50I yule Ijuwl 

— ir^ ofScand, j5l. geoldei Lay. yolnight MH — (juul) Bch. 
fgollprt tfell fell fbl 

— fe<il (u), feoUen (voile) pi Ixiy. uolle : belle Horn, iuel KS. Til Ay, 
feU: tell Cif. fel (u) PP/. '^\,eCh. 

hdoldprt theld held held 

— e(a), (e)o Ld. eo, u (e) Lay. eo AR. e K8. i Ay. e (u) PPL 
e(i) North, e GE. helde pi : schelde RBC. u JTarZ. beeld CA— 
(hild) hai'bare Cp. 

Beo]>aii 8e}>en seethe B^tS 

for-leoean t-le(o)8enn lose luws 

— eo (ea) Lay. eo AR. he l(e)oet, lust ON. ie JKiS. ye j4y. e North., 
Awi., Ch. eo Harl. e, o WiW. lo(y)8e:ho8e ih TM—lose Td. 
looeeth Ck. 

1780 freosan fresen freese triiz 

(ii) G 

fleoB flea fleece flije 

fheoaan fhesen eneese enUa 

— fneeeth (sneseth) Ch — (niit) Sm. 



FIRST WOED-LIST. 357 

o9o8aii toheeenn choose tjuws 

— oeaen Li, oo Lay., AB, \e, ye Ay. e North,, GE, Ch, Wid. 
al9Q u — u Ch. (yy) Q. (t/auz) Mg, 

prOost tpre(o)Bt priest pr^st 

— precte, i (ei) : nesie superl. CM, pniest Harl, e, i, y Aud. 

1785 brdost tbre(o)st breast brest 

— (e)o Lay. eo AB, ye Ay,, ee Wid, brest : feet ( « y), lest 'desire/ 
fest 'feMt/ biist (e) CAr-(brespl8eBt) Jn, 

Sow ew yew juw 

al*o iw (yy) £^01. 

Sow tjuw you juw 

iu(i)h Du, eow, iu Bu. — (e)ou, oeu, jeow (30U, on) Lay. ou AR, ^ew 
Prod, ou, eu OiS^. yw iiT^. giu (yow):IIieAU CM, yow:now i/r. 
gu (7^. ou Harl, yow ^//P. 30W : knowe inf Aud, jow : now Ch 
— yw,yoHVg, (juu, jou) G^. (jiu,jou) Cp. (juu) lid. 

Sower tjure your juer 

iuer Du, e(o)wer Bu, — eoar, euwer, ower Horn, ower AB, yure K8, 

S'ur (yur) C3f. guro (rJE, o(u)r ffarl. gour i4ii(/. youre Ch—yv/r 
Yg, (juur) G^. 0'"^ <^P- 

hSow sn thewe hue l^uw 

hiw W8. 

1790 hrSowan treowenn rue truw 

— rew : new, thou rues, rufully TM — (ryy) Fg, G, 

]>rSow prt ]>r6U threw ]»ruw 

sSowan sewen sew sou 

ainuid Ep, siuwid Cp. i» y W8. siuie9 Du, liowes Bu, — seouwen 
AS, sewe (seu) CM. sowed (swed, sewed) Ch — (seu) Bt, (soou) G, 
(siu, soo) Ld. (soo) jBe&, Sh, 

ftower tfowwerr four for 

feor Du, — fewer Ld, f(e)o(u)wer, f(e)our Lay, iio(a)r AB. aoDr Ay, 
four, faur North, foure TJIf. four i/aW. fawre AllP — (oou) O. 
(ou) Cp. (00) lid. 

ftowertSne fourtene fourteen fot^n 
(oou) G, 

1795 fSowertSne-niht fourtenight fortnight fotnait 

i^owertig tfowwertJg forty foti 

— feowerti Lay, fowrti Jul, fnrti £6^. uourti Ay. foorti (forti) CM. 
f*urty MB, AUP. fourti rjf— fourtye Td, (o) 0. (00) Ld, 

nSowe tne(o)we new x^uw 

io, eo VP, Bu, niwe Du,, WS-^neuum Ld. ncowe Zoy., AB. newe 
ir<. nea Norths — (nyy) G, (nyy, neu, niu) W, (niu) J8cA, iSA. 
(nuu) Fr. 

cSowan tohewwenn ohew tjuw 

(tjiu) Cp, (00, oou, iu, 00) Jn. (iu, 00) Ld. (uu, 00) Sh. 

— (00) vg. 

crSow prt creu crew kruw 

1800 olSowe dewe dew kluw • 



'^ 



58 HISTORY OF ENGLISH SOUNDS. 



endow prt ton6(o)w knew x^uw 

— ^knew Ay., North, bicno AUP, knoghe : enugbe Tilf. kne^ Wicl, 
— (yy) G. (knyy) W. (nhiu) Cp. 

grSow prt greu grew gruw 

trSowe ttrowwe true truw 

tre(o)we VP. eo, ea, trewe Ra. ie, y WS — ^treo(u)we (trewe) Lay, 
trewe (eo) AB. triwe RGl, trewe Ay, irewe (trea) : neu» treali 
(truli) CM, trewe OE, Harl. trwe AllP, trew, truli Aud, trew : 
blew 'blue/ truly TM. tr(e)uli F^ici. trewe :newe CAr— truely Td. 
(yy)0. {iu) Cp.Beh, (}i}i)Fr,Sh. 

trSowian ttrowwenn trow ttrau 

— ^ti(e)owen (trouen) Lay, (trowen) AB, tru (trau) : nu CM, trow 
PC. tro:d6 Hv. trowen, e OJS, trawen, tryjeiyje, trow AllP. 
trowe:bowe («= boga), )>ey trowd (trod): God BBC, traw:draw, 
trow : now TM, trowe : growe ( = 5w), blowe ( « &w) Ch — tro IlVg, 8b, 
(oou) O. 

1805 brdowan brewen brew bruw 

(yy) 6m, W, 

bleow prt bleu blew bluw 

* blew,' * bloomed.' 

*hreowJ> rexipe ruth truwj> 

— reou^e (reuj>e) Lay, reou9e AB. Teu)»e Ay, eu North, ew GE, 
ou Harl, »u AllP, routhe : slouthe, eu (au) Ch — (yy) O, («) Bch, 
(uu) 8h, 

f.4^»Y. +fwv««i.** /truth truw]> 

treowp Ttrowwpe < , . ' 

'^ ^ (troth ttrou}> 

— treuthe, 60 Ld, treowtJe, treuwe}« (tpeuj>e) Lay. trou9e AB, 
treu>e BGl, Ay. troube CM, trou>e, au, ew MB. trawthe : cdawthe, 
(o)u TM, trev/iSe OR tren>e, ou Harl. traw)»e AUP. trowth, eu 
^ud.— trueth, eu, truwth, ou Td, (yy, u !) Q. (bitroj)) betroth Pr. 
(uu) Bch, Sh, 

leof leef Uef tiyf 

— luef HaH. 

1 810 ISof xnann lemman lenunan tlemen 

— leofmon Lay., AB, lemman Ay., Ch. 
\>eot J>ef thief J>yf 

—Peha Ld. J>yef. ]>ief ^y.— (ii) G. 

oleofajL eleven cleave kl^v 
(ii) Sm, (ee) G. 

dSofol tdefell devU devl 

diobul, diowl, diwl, diul Dm.— dyeuel Ay. devel North, diuel Bett, 
deviUe, dewiUe, dwylle TJf— devyU, dyvell Td, devel Ck, (diivU) 
5«i. (devl)Cp. (divl, dU, del) J,». {devl) Bch, Sh, 

betwdonan tbitwenenn between bitwjjn 

— betwenen Ld, bitweonen, u, bituei5en (bitwine) Lay, betuene K8. 

— (ii) G, 

1815 *gebSonptc Iben l>een bi(j)n 

hynTd, (n) 0, (i) Jn, Ld. 

fSond tfend fiend 4jnd 

— u(y)end Ay. fynd : kynd adj Awi,—f[ynd HVg. (ii) W, (i) J. 



FIRST WORD-LIST. 359 

&6ond tfrend firiend frend 

— ur(y)end Ap. frendihend CM. ey TM. y Aud. ee Wiel., Ch — 
ffrynd HVg. fwndly Td. (i) O. (U) £L (U. i. e) Jn. (fifenU, 
frenjip) Jn, (friind) EO. (e) Bch, Sh, 

Bfioc 'Ne(o)o 0lok sik 

e Du,, Bu, eOy e, 8By u (e*) Lap. see, ye rike, Bi(o)kne8te ^i{. sik, lik 
Kt Bek : chek MH, soke : speke TM, seke ^«ul. fiijk, nek WicL 
■eeke |y^ : seke vb, dk : phisik CA — rioke, e Td. 

XTOoiVik tmeoo meek ' m^jk 

1820 pMi t)>e(o)s pi thigh pal 

— ))eh, )>ih Lap, )wo:beo OiNT. >o:to be JBBC. thee: me 7j|f. thi 
Primpt. the C^— ()Ah) JB/^ 

leoht an, adj tlihht sb, adj light lait 

<bn^ht(neMV e VP, Du, 90, e, i Ru.—]ihi adj Ld. liht Lap, AB. 
liht : night ON, li^t Ap. lyht : Bjht, lyt : Hyt ( - site])) ifaW. 

leoht tlihht Ught lait 

Mevis.' e2>tt. i J2«.— leht» i Xay. UhtiiR. Uat ^y.— (Uht, loit) iSfm. 

Idogan tlejhenn lie lai 

'mentiri/ e, i VP, e Du. i Bu. — ^li^en (e) Lap. li^en (lihen) AB. 
leghen K8. \{Y)e^e Ap. ]»u leies (lighes) CM, lye PC. +ly TAf. 
ley : fley (* flew^p/) BBC. legen (?J&. flye Ifor^. fHe C^Ui) G. 

fldogan tflejhenn fly flai 

e VP, Du.-'fieon Lap. yleon AB. ▼!!, he Tlij)) ^y. fleie (fli) CAf. 
fly : ly ( - licg»«) TM, fleye : heye (high adv) BBC. flee : free Ch. 

i825fldoge8f flie fly flai 

— fle(o)5e (fleie) £ay. Tli^e (flehe) AB, y\e$e Ap. flege QB, fle-wing 
TM, fle3e (flie) Wicl. fme Ch, 

g6ogu)> sf S^®]'® youth i\iwp 

iugafSf gogny VP — ^gajeOe Lap. juweOe AB. ye^e]^ A^, youthe: 
monthe MH, yoQthe : nowthe ( «* nO )»&) Ch-^uu) Bt, (yy) G, 
(ia)€p. (»)/ii. 

tdo(go)]>ian ti]>en tithe taiS 

teigf^ Du. teg>igan Ru. teotSi(g)iAn 2F2J^-teo>egeii AB, teGien 
Horn, te>en (i) PP/. i Promji^.— (tai» U. 

]>rdotAne i-]>rittene thirteen }>eetUn 

Bodotan tahetenn shoot Juwt 

Lap., AB, e QE, Wid,, Prompt, iCh. also u. shotee TM, 

flUt 

but 

lijd 

wtid 

hips 

step- 
kr«p 

d«p 



1830 flAote 


flete 


fleet 


fliute 'nitis* Erfglr-^ Lap, 




bdot pit 


beet 


beat 


— eo -41?— (e) 8m 


\, 




hrdod 


red 


reed 


wdod 


wed 


weed 


heope 


hepe 


hips 


'dog.roee.' 






1835 8tdop(fMer) 


step- 


step- 


ortopan 


erepen 


creep 


(ii)a. 






ddop 


tde(o)p 


deep 


(ii) G, 







36o 



HISTORY OF ENGLISH SOUNDS, 



I. 

U +bl \yy bai 

only adv — be, bi prp Ld. be, bie prp KS. be, by prp Ay, be prp 
^fk^.— (bi) 8m, (bei) O, 

iren iirenn iron aien 

isern Ep. i8e(r)n, iren VP — ^imene adj 0. ben Jy. — ^yeron Td, (oieni) 
G. (aidm, em) Jn, 

1840 acir sf shire shire Jaier, -Jler 

(wuBterfiiOa (Si\i) Cp, Jn, EO, Ld, Bck. (Joir) 8k. {i\)weak 

form, 

scir-gerdlfa sohirreve sheriff 

spir spire spire 

hreodes spir Leechd, 
V9VC wir wire 

hwil 6f while while 

{while 
whilst 
— aUo wllenn 0, )» while Lay, )>eo hwule AR, ^e wale ]»e RGl, 
whiles PC. wile, quilee «a( G^i?— hwill Td. (hweU, hweUs, hwlls) 
8m. 



1845 ]>ft-hwile-]>e thwil 



Jerif 
spaier 

waier 
whail 
whail 
whailst 



filsf file 


file 


flEdl 


eo W8. 






mil sf mile 


mile 


mail 


pU — 


pile 


pail 


pilas ' htiin of plants ' Leeehd, 






pn pil 


pile 


paa 


'javelin," stake.' 






1850 Ujje tU]>e 


Uthe 


laitS 


* gentle.* 






Ei]>e si)>e 


scythe 


salts 


wii]>an wri]>eii 


writhe 


raits 


(rai)») Ld. 






{ti}>indl snpl tti]>ennde 


tidings 


taldl 



— tiCinde, tidende, tidinge Lay, tiiSinge AR, tithand North, tiding 
GE. tydand : Scotland, tydinges : offi^ges RBC, 

bU]>e tbU]>e blithe tblaitS 
(bleiW Cp, 

1855 is is ice ais 

a-risan trisenn rise rais 

yniB "twis wise waiB 

(weis) Sm, (woiz) G. 

reht-wxs trihhtwis righteous traitjes 

— rihtwis Lay., AR, ri^twis Wid. — rightewesness Td. (reixteus) G. 
(raitiuB, raiteus) Jn, (raitjim) Ld, 

wise sf twise wise wais 



i 



FIRST WORD-LIST. 36 1 

1860 ongxiallo tgriMdij grisly tgrisli 

also y ( B 6 ?) — gTulich Lay,, AB, Ay. gru<(e)li PC. aUo e — (groizli) G. 

Crist tcrist Christ kraist 
kreiat BVg. 

oristen-ddm torisstenndom Christendom krisndem 

eristnian forisstneim christen krisn 

*Cristes-mflB6B6 oristesmesse Christmas krismes 

— Ay, oristemMse Ch — (krisnues) Jn. 

1865 grist grist grist grist 

wis'ddm twissdom wisdom wisdem 
(wiizdum) BU. (wizdum) 0. (s) Ld. 

hiwa bine hind thaind 

' inmate of fiunily * — ^hinen pi Lay. nfiffpl hiwena (?). 

]yxiwa t)>ri5{5)e6s thrice t}>rais 

0ri(g)ft Du. fftige, ]Nriuwa, Jnriowa Bu. — >rie(n), Jrreie, Jn^oien (^ribn) 
Lay, ynet AB, Ay. Jfrise : wise CM, thries CL 

splwan spewen spew spjuw 

n<d in Angl. — Bpeowen ; Bpi imper. AB. spayd pte TM. 

1870 tiwes-dsg tiwesdai tuesday tjuwsdi 
(tiaxdi) B^. (tjousdee) 8h. 

twiwa ttwijess twice twais 

twiga Du., Bu. tawa Or. — twiges Ld, tw(e)ieii (twie) Lay. twien, 
twie (twies) AB- tayes Ay. twie GE. 



iflg 


tivl 


Ivy 


aivi 


lif 


tlif 


life 


laif 


on life 


on live 


aUve 


elaiv 


1875 XpMBak. 


Jniven 


thrive 


J)raiv 


soiifSeui 


tshrifenn 


shrive 


tjraiv 


♦stif 


stif 


stiff 


stif 


KtTfian'itiffeii.' 


»U>'Btiff'~iiC3f 


• 




wif 


+wif 


wife 


waif 


wlf-maon 


1 Iwimmann , 


woman 


wumei 



wimmaii late — wimman Ld. wifmon, wiin(m)oii Lay. wummon AB. 
wyman K8, wyfmanne d. Ay. wimman, wom(m)an CM. woman, 
weman PC, wymmon, wommon Harl. wemon Aud. — (wuman) G. 
(wuuman) Bi. (w«m»n) Pr. (mnsn) Jn, (wmiiaen) EG* («) not 
to obscure oi in brother etc Ld. (wvmin) Bch. (womon) Sh — vg (umon). 

1880 wif-mfnn twifinenn women wimln 

— wimmen Ld. wnmmen AB. wyfmen Ay. wimmen MS. wymmen 
Harl. wommen C^^wemen Td. (wimen, wiimen) G. (wiimen) 
Pr. (wtmen) Cp. (wimin) Bch, Sh. 

at tflf five fS&iv 

fife pi — fyf, fyye CA— (fipens) fivepenoe Jn — older and vg (fipona). 

nfta tfifte fifth fif]> 

— fiahe (fyfte) CA— fyfte Trf. (fia) (?. 



362 



HISTOBT OF ENGLISH SOUNDS. 



tooif 


onif 


knife 


naif 


lat£;frScand.kmfO), 






diiflux 


tdrifenn 


drive 


draiv 


1885 fifUg 


fflflftij 


fifty 


fifti 


line 


line 


Une 


lain 


linen adj 


linen 


linen 


linln 


— ^linen AB, Ay. 


liiie(ii) Wid, 


linnen Lap,, CM, 


PPL 


Jnn 


t]>i{n) 


thine, thy 


tt!ai(n) 


Bwin 


tswin 


swine 


swain 


1890 Boinan 


tshinenn 


Bhine 


ain 


Borin 


Bohrin 


shrine 


rain 

1 


win 


twin 


wine 


wain 


hwinan 


whinen 


whine 


whain 


min 


tml(n) 


my, mine 


mai(n) 


1895 twin 


twin 


twine 


twain 


tdwinan 


dwinen 


dwindle 


dwindl 


(dwinl) Jn. 


* 






pinian 


tpinenn 


pine 


pain 


•torture.' 








pin-trdo 


pine 


pine 


pain 


Un-B«d 


linsed 


linseed 


llnsfld 


1900 rim 6n 


trime 


rhyme 


raim 


hrim 


rim 


rime 


traim 


lim 


film 


lime 


laim 


slim 


8lim 


slime 


slaim 


tima 


itime 


time 


taim 


gelio 


tUc 


like 


laik 


(i) Sm. (ai) G. 






1905 tidcan 


tsikenn 


Biffh 


sai 



— liken, lichen Lay. nken Ch — (sih, aeih) Stn. {sai, teiy) Jn, (lai^) 
Ld, (sei), heUer (idi» Bch. (loih) Sh, 

tsnioan sniken sneak sn^k 

— Bhrike Jraik 



Bono 



striken 



strike 



straik 



strican 

* glide.' 

( dik dyke daik 

t dioh ditoh ditj 

* trench '•— dic(h) Lay. diche pi : riche at^ Moral Ode. dich Ay. dik 
GE. dyche, dyke : 1 jke TM (Ul^' trenoh '— deitoyi ' dhchei ' Sb. 



dio sm 



1910 pio 


Pik 


pike 


paik 


a-lihtan 


aliht^n 


alight 


elait 


higian 


thijhenn 


hie 


thai 


— hih »& 0, 








Btig-weard 


Btiward 


steward 


stjue 


—Ld, AR, . 


Prompt. 











FIRST WORD-LIST. 


36^ 




■tig-rftp 


stirop 


stirrup 


stirep 


1915 


snite 

— Wid, mipe 


snite 

Prompt. 


snipe 


snaip 




■mitan 


amiten 


smite 


tsmait 




'smeur.' 










t»t-witan 


atwiten 


twit 


ttwit 




' reproAch ' §b edwit—edwit, asdwii 


Lay, edwit AB, 


atwiten Lay,, Ay 




etwiten, edwiteii AB. 








writan 


i'writexixi 


write 


rait 




hwit 


whit 


white 


whait 


Ipso 


mite 


mite 


mite 


mait 




bitan 


tbitenn 


bite 


bait 




idol 


tideU 


idle 


aidl 




hi[gi)d Bfe 


hide 


hide 


thaid 




'meMoreof land.' 








ridan 


tridenn 


ride 


raid 


1925 


Bide 


tside 


side 


said 




aUdan 


aliden 


sUde 


alaid 




atiidaa 


striden 


stride 


straid 




wid 


twid 


wide 


waid 




cidan 


ohiden 


chide 


ttjaid 


1930 


l^dan 


gliden 


glide 


glaid 




tidsf 


ttid 


tide 


taid 




bidan 


biden 


bide 


•tbaid 




bridel 


bridel 


bridle 


braidl 




— faildledd 0. 










ripe 


ripe 


ripe 


raip 


1935 wipian 


wipen 


wipe 


waip 




ffripaa 


gripen 


gripe 


graip 




pipe 


pipe 


pipe 


paip 



ha thu how hau 

— hn Lay, li(w)a AR. hu, qahu OE, hu, wu Best, it^, of hwj— 
how BVg, 



jm 


+Jm 


thou 


+«au 


1940 na 


iiiu 


now 


nau 


oQ 


cm 


cow 


kau 


Ibaisk 


biisken 


busk 


tbask 


'prepare 


oneielf' — alao 0. 






brasf 


bruwe 


brow 


brau 


fire 


ture 


our 


auer 



owr, our HVg, (our) Q, (ouer) Bt, 



364 



HISTORY OF ENGLISH SOUNDS. 



1945 t1'^^^''^~^^fi^ Jnirsdai Thursday ])eeBdi 

yftr/^m Scand. ^r ; OE ^onreads^ — JtimresdiBi Lay. )>ur8dei AR, 



BtU* 


8ur 


sour 


sauer 


sower Ck. 








acfir 


Bohur 


shower 


Jauer 


bfir 


bur 


bower 


tbauer 


(hour) G. 








fOr-lang 


fUrloDg 


fUrlong 


feelox| 


■■ furh — forloDge 


AUP. furlong 


(forlonge) Ch. 




1950 tkle 


ule 


owl 


aul 


mi 


tftil 


foul 


&ul 


(foul) Sm, 








8li)> 


tsuj> 


south 


8au]> 


8li]>eme 


8ou]>6me 


southern 


saSen 


mil)) 


tmuji 


mouth 


mau]> 


1956 <5<i]>« 


tcu])e 


could 


kud 



— ouSe Lay., AR, couthe : raouthe MH. tcowth, cowde TAT. kowtbe 
Aud. koutbe : Dertemonthe, koude:loade Ch — coulde, culde Td, 
(kould) Sm. (kuuld) & etc, Cp. (kould) Pr. (kuud) Jn, Beh, 
(kuuld, kuud, kud) Ld. (kud) Sk, 

•an-ctp tunncu]) uncouth ankuw]> 
(uu) Cp. (y) Jn. 

fLs fuss us as 

— ou8 Ay, (o)u8 HarL ous A UP, us : precius Ch — ws HVff, 

htis thus house haus 

lus lus louse laus 

i960 )>tLsend sn t]>u8ennde thousand ]>au2nd 

— ^usen Ijd, ]>u8und, ]>u8end, ]>u8eDde pi Ltiy. )x>usond Ay, ^usanrl 
North, — (]>ouzand) O, (jiouzn) Cp. (Jwuzond) EO, (]>ou28eDd) 
Fr, 



mQs 

drfksian 
Grein, 

^h1is-)>ing 
' meeting * — Lay. 



mus 



husting 



mouse 
drowsy 

hustings 



maus 
drauai 

hastiiji 



hfUi-wIf 



huswlf 



{housewife hauswaif 

hussi^ thussy haBi(f) 

— ^husewif AR. houswif PPl, huswif Prompt. — (hvzif) Mg, (hvzii, 
hwi, hvsi) Jn, (hvaiy) Ld. (hvzif) Beh, (hvzwif) 8h. 

1965 dtUit tdusst dust dast 

— u (ou) Lay. ou Ay. 

hfUi-bonda husbonde husband hasbend 

from Scand, hOsbOndi 'house-roaster.' bonda, bunda Laws ^pr, — 
Lay.^ Prompt, housbonde (0) Ch, husbande : sland CM — husbande 
Td. 

scQfe sohuven shove Jav 

— scuyen (seve) Lay. schouve Ch — («) Ju, 



FIRST WOHD-LIST. 



365 



tgrlUiEi vb grovelinge grovel 

'crouch' — alio u — (gmvlii)) EO, (grovlii)) Beh, Sh, 



{hftre-htLne 



horehune •< ^ > hound 

(hoar j 



groTl 



hor(h)auiid 



1970 ton 
dfLn Bf 
of-dtine adv 



ttun town taun 

tdun down daun 

tdiin down daun 

aUo dunestigan VP — dun Ld, dannwarrd 0, douni donwArd BQl, PC 
—down MVff, 



JdOn 
'feathen.' 

pUnian 
—Wiel 

1975 tbiiin ptc 
'rcftdy.' 

brOn 

rflzn 
roume TtJ, 

)>tkina 



doun 

po(u)nen 
tbun 



down 



pound 



bound 



daun 



paund 



baund 



brun brown braun 

trum room rum 

(qu) BU. 

)>ume thumb ]>am 

— ^ume Ay, thoumbe Prompt thombc (thome) Ch. 
plfbne ploume plum plam 

1980 stioan suken sudk aak 

— zoake Ay, ■owke:crowke Ch, oo WicL, TM, aokin Prompt.^ 
swking HVff. soukiingea Ck, 

brQcan tbrukenn brook tbruk 

* enjoy * — ^bn]k(i)en Lay, on CA— (uo) G, 

rtkh truhh rough raf 

—nigh North., TM. ru : Esaa GE. rough Ch., TTic/.— (mf) WtU. 

thunig-afkge hunisuoole honeyauokle hanliwlrl 
— ^honifloole Prompt, 

{kUga — oow kau 

' subdue/ 

1985 drOga]) tdruhh]>e drought draut, drot 

— draj^e Ay. drugte &E. dronghbe, dro(u)ghte PPl. dr^ghte (ou) 
CA— (dreut) Mg, (eu, 00, 00) Jn, (dr»u>t) Ld. (ou) Beh, 8h. 
(dx9uti) Behf (drouti) 6h droughty. 

bfLgan tbujhenn bow bau 
(bou) 8m, O, (buu) BU. (bou) * torqueo' Cp. (bau) Ld. 

lit tilt out aut 

owt HVff. 

ymb-titan abuten about ebaut 

aho onbutan (about) G, (ebeut) Cp. (bout) Jh. 

titerre cp utter utter ater 

—utter JRy PC, Ch. outtreete (outreete) Ch, 

1990 wi]>-fltan twi)>]mtenn without wltSaut 

ItLtaa tldtenn lout laut 

ciat tdut olout klaut 



366 



HISTORY OP ENGLISH SOUNDS. 



btitaa tbutenn but bat 

— ^bute Ld. also btLt(t), butt 0. beute, buten, botr(ii) Lay, bDte(n) 

AB, bote Ay, bot, a North, boute prp, bote ctmj Earl, — butt Td» 

priit prud proud praud 

— prut (ou) Lay, fprout RGl, prud AR. proud Ay, fprud QE, 
proud Ch, 



1995 lider udder udder 

— uddir, iddir Prompt. 

hlM tlhude ay 

Borfld tehrud 

orudan oruden 

* press.' 

cltid tclud 

*rock.' 

2000 lip tupp 

upp hy if^fl, of uppan — oppe, ]>erop Ay, 

Blipan Boupen sup 

— Prompt, 

BtQ.pian etupen stoop 

— u (ou) Lay, ou Ay,, PC, Ch — (uu) Cp, 

tdrapa drupen droop 

* be dejected ' — drupand (uu) CM. ou Ch, 



doud 



up 



ader 



loud 


laud 


shroud 


Jraud 


orowd 


kraud 



klaud 



ap 



sap 



stuwp 



druwp 



takf en side sky skai 

' cloud.* OE woloen — a Uo skewes pi, 
2005 hw7 twhi why whai 

0^ pi kyn kine tkain 

— ^ken Ay, ky(e) North, kyen Wid, keen (kyn) Ch. 

tbycgaa tbiggenn buy bai 

imper, byge — ^buggen (i), bu^e imper. Lay, beggen, he bay^ Ay. f^ie 
North., TM. by(y)6 AUP, bigen 0£. abc^je : l^gge, t^^bye, beye : 
tweye Ch — (bei) om, O. 

hyran hlren hire haier 

f;^ tflr fire fiaier 

— i Ld, u Lay,, AR. fuyr RGl. f(u)ir PPl uer Ay, i North. 
Iy(e)re : myre TM. fi(g)er, fir QE u Barl, foyre : hit wSre, fouyre, 
fuyre, feyre Aud. fyer Wiel, fyr Ch — (foier) &, 



2010 ^m^r sf mire mire 
' swamp ' — alio u, ie. 

fyian tfilenn (de)flle 

b^le bile boil 

'ulcer* — also u, ui, e — (ei) 8m — vy (bail). 

f71]> sf fll]>e filth 

oy^}>o ki]>])e kith 

' home/ ' firiends.* 



maier 

difkil 
boil 

fil}> 
tWj) 



FIEST WOED-LIST. 367 

20151^8 pi Ub lioe lais 

(©i) or (ii) JSen Jonmm, (ii) harhare Cp, (u)/rOfli ICt *l68. 

mfn pi mis mioe mala 

(91) or (ii) Ben Jomon, (ii) harhare Cp, (ii) from Kt *m6t. 

wysoan wiflohen wish wij 

— ^wenen 4^. wusoLin, i Prompt, — ^wish HVg, 
tj^rysta ]>rasten thrust Jnraat 

-^ Ay. i (?J?, Prompt i, t Wiel thrMte:lfigte (i, u) <dMired' CA. 

9st sf fist fist fist 

Q Lajf., AR, AllP. i Prompt, (at (!) : brait ( »6o), bett adj Ch. 



2020 hff b£ hive 


hive 


halT 


dyflin diven 
—Q AR, t Marg. i PP/. 
br^no brina 


dive 
brine 


dalT 
brain 


]>^el thimbU 

* thumlMtaU '—Prompt. 
dryge tdri55e 


thimble 
dry 


]Hmbl 
drai 



— dniie AR. dri North, dry^e, druye AllP* drie Wicl. dreye : weye 
(drie) Ch. 

2025 Vtel tlitell Uttle UU 

— f>< little 0. Inttel (t) J^oy. lutthle pi, lutie Ifom. lotd AR. litel, 
p2 HtUe ily. lutel Harl. lyttil, ly ty iiiMi. litel, lite : vimte Ch. (liU) 
• pAryas ' (liitl) ' valde pamis ' G^— (lait) Lyte. older (lijtl). 

oyta kite kite kait 

-^Ajf. j{e)Ch. 
pryte pride pride praid 

— ptQie, vnde I^ay. prate ROl. prnde AR. pnAeAy. pruyde PP/. 



pride GE. 








hfdd 


hide 


hide 


haid 


hydan 


thidenn 


hide 


haid 


2030 bryd 


tbrid 


bride 


braid 


hydde prt 


thidde 


hid 


hid 


drypan 


drippin 


drip 


drip 



/ WS ; *■ 8 (V. — Prompt, alto drepen. 



O. 

scd tsho shoe Juw 

ahues Td. (uu) Pg. 

t5 adv tto too tuw 

2035 ^ PT "^f to to tu, te 

->io, U Lay. to, aor te ulR. to (to) CJf— to, tw BVy. (ta, to) O. 
{UM)Ld. 

*mi-t5weard — untoward tantoued 

(teoierd) toward Pr. (o) not (o) tii toward Ld. 



368 HISTORY OF ENGLISH SOUNDS. 

d5 yb tdon do duw 

(dua) 8m. (du, duuBt, daii)) G, (duu) reetius (doo) W. (duu) Pr etc, 

dra oor ore oer 

— also ore (?). 

hora hore whore hor 

(Luur) Sm, Pr etc, 8h. {hoor, whoor) Ld, (hoor) EO, Beh — older 

(huer). 

, .. ( wwaxe ffsweer 

2040 swdr prt swor < < 

{ swore I Bwor 

— o Laif.f GE, Euor Ay. Bwar (o) AR. aware North. — (00) G. (fonuur) 
Jn. (soor, Bwoor) id, 

fldr sfm tflor floor flor 
(fluuer) tometimet Jn, (00) Ld. 

m5r mor moor miier 
(moor) EO, Ld, JSch, 8h. 

stol stol stool stuwl 

8c51 sf scole school skuwl 

Aifcgl. ucoiu eWS. 

2045 ool col cool kuwl 

tol tol tool tuwl 

pdl pol pool puwl 

d]>3r idperr other aCer 

— d^replO. ootherCA — {u,o) 6m, G. (aenstJar, n*Sor) J«. (p)not{o)Ld. 

xd|>or ro]>er rudder rader 

2050 b5]> iadp sooth tsuw]> 
(forsv]') better (forsuu))) J*. 

waaope adv i-Bnie]>e smooth smuwtS 

aUo a* adj inetead of BmS]>e — e Alt, "fPC. o Prompt., Ch. 
td]> to]> tooth tuw)> 

tdGB]> vb tdo)> doth tda^ 

— dooth Ch. doithe Tif— dwth ITF^. (u) 0^. (uu) /». 

]:bd]> sf tbo])e booth buw]> 

JEScajki. ; 01 ba> (buu«) JBll. 

2055 br5)>or tbro]>8rr brother braSer 

oCh. (u) a (bri*«erhtid) Cp. 

osle osel oiuel uwbI 

Upus los loose(n) lttWB(n) 

Iguana * get looee ' — (]» lowae aton, louaae) AB. looe ' aolntua * Prompt, 
laua (louse, loa), looaigooa Ch — lowaen, loose r6 Td. looua, loua, loos 
C^. (uu) 8m. 

wda wos ooze uwb 

goa gos goose guws 

— guoe Ay. 

2060 gos-hafoc goshauk goshawk goshok 

b5sm tbosemm bosom buzem 

(b«z9in) Jn. (bvzem) Fr. (bozem) Beh. (buuzom) Sh — vff (basim). 

hrost — roost ruwst 



FIRST WOBD-LIST. 



;69 



rdwan 


rowen 


row 


(roou) Bll 






hldwan 


lowen 


low 


(ou) Sb, 






Btdwian 


stowen 


stow 


fldwan 
grdwan 


i*flow0xui 


flow 


growen 


grow 


(oou) 0, 






fl^dwan 


tfirlowenn 


fl^OW 


bl5wan 


blowon 


blow 


* bloom.' 






hdf 


hof 


hoof 


— hufe PC— (huuv) Sm, 




hdfprt 


thof(f) 


hove 



fditor tfbsstreim vb foster foster 

mdste prt fmosste must mast 

— o Xay., AR, CK u Horn,, GE, TJtf— muste Td. 
2065 bloBtm(e) tblosstme blossom blosem 

- — ^blostme AS, bloflme Mttrg., Ch, bloeme, blossuni Prompt.— bloawm 

tSwef oof woof wuwf 

(w«f), Mi^ (uuf) Jn. w it^, of wefiin. 

rou 

lou 

stou 
2070 fldwan tflowexm flow flou 

' grou 

glou 
blott 

hu(w)f 

2075 hdfprt thof(f) hove thouv 

—hof, n, M (eo), hefde Lay. hef AR. hof North, baf : yaf, haaf Ch. 

be-hdfian tbihofenn behove tbihouv 

— bchoued : loued CM. byhmfeilufe PC— (bibuuv) C, C/>, EO, Sh. 
(bihvv) Pr — older (uw). 

wudu-rdfo wuderove woodruff wudraf 

hrfif tr(h)of roof ruwf 

gldfsf glove glove glav 

(gluv) &. 

2080 proflan provien prove pruwv 

— preoven AR. provi Ay. provea, e PFl. ir^. of F)r — (uu) Bi, Cp. 
(«) Pr. («, uu) EO^older (a). 

w6tte tsoffte soft soft 

adv of tffifte, but dUo a<?;— aoofte Td. (o) Sm, O. 

sdna tson(e) 

spon Bpon 

'chip' (uu)a. 

n5n non 

2085 mdna mone 

m5]ia]> mone]> 

(mun])) O, 

mdnan-dAg mo]ie(n)dai Monday mandi 

— munendai OE — (mundai) Bi. (muundee) Jn, (mvndi) Bch. (mvn- 
dee)^ 

geddn pic tdon done dan 

—don (u) CAf— dynn, dy wn BVg. done Td. doon Ck. (u) Q. 

Bb 



soon 


suwn 


spoon 


spu(w)n 


noon 


nuwn 


moon 


muwn 


month 


mao]> 



370 



HI8T0EY OF ENGLISH SOUNDS. 



boon tbawn 

OE b6n tf—aUo bene O. bone Xay., A'Rn 



loom 



luwm 



tbdn sf tbone 

'request'; general^ ben. 
bone : dene adj Horn. 

2090 15ma lome 

'tool/ 

cdm toomxn. came keim 

— oom LA, odmenn pi 0. com Lay,, AR. oom:-d5m ROL kam 
KS. oom Ay, oom, cun : Adam CM, ciun : Adam &^. cam : Dar- 
daniam, oom : dSm RBC, oom : dome TM, com J2ZP. come : dome 
Aud. cazfi : ram, ooom : -dom Ch — camCe) Td, (bikaam) Q, 



goma gome gum 

glom — gloom 

ddm tdom doom 

209g -dom t-dom -dom. 

dom PC, -dam QE, ARP, Aud., fRBC. 

•dom : coom, -dam Ch dome, -domm Td. 



gam 

glttwm 

duwm. 

-dem 
-dam, -doom WieL 

bru(w)m 

"fblawm 

bluwm 

hiik 

rule 

Ink 

(un) 8m etc, («) belter (u) /». 

Idea imper. tloke, loc lo tlou 

lochwaet ' whateyer ' etc late — ^loa, leo (lo), la Lay, lo (low) AR. lo : do 
C3f— <hallaw) hollo. 



brom 




brom 


broom 


(uu) BU. 






bloma 




— 


bloom 


' maoB of metal.* 






tbldmi 




tblome 


bloom 


hoc 




hok 


hook 


2100 hroo 




rok 


rook 


looian 




tlokenn 


look 



Bohok 



shook 



Juk 



wok 


woke 


woiik 


cok 


cook 


kuk 


toroc 


crook 


kruk 


ttoo 


took 


tuk 



scoc prt 

(nu) G. 

on-woo 

2105 000 

(un) Q. 

tkrok 

ttoc prt 

(«), better (u) Jn, 

boc tboc 

(uu) G, Cp, Bch, Sh. 

broc brok 

2 1 to hoh hoh 

(hof) Dyehe. (hok) 8h, 

toh touh 

(tou, touh) Sm, (t«0 W, (too) Jn, (t«f) EO, Bch, 8h, 

]>ohte prt t]>ohhte thought ]x>t 

(jpowhi) Bll, {pQOVLxt)0. (t>oot) Pr, J&a (p) Beh. (po)Ld,Sk. 

Bohte prt tBohhte sought sot 
(louht) Sm. (bifoot) Jn. (00) Ld, 



book 

brook 
hough 

tough 



buk 

bruk 
hok 

taf 



FIRST WOED-LIST. 37 1 

brohte prt tbrohhte brought brot 

(brouxt) 0, (broot) Pr, Jh, (o) Beh, (oo) Ld, Sh, 

21 15 Bldg Blouh Blough slau 

'deyinm' Wgt, 'qaoddftm oonoaTum' Bede — (00) Ld, (alvf) Bek, 
{aha) 8h, 

Bl5g prt ttdoh Blew tsluw 

— pi dogfaenn 0, doh Lmf, ilonh AB, alobiinoh ME. doff, slug 
GE. slow : how adv BBC, BI005. slew Wid. ilough, dow CA-~dae 
Td, (yy) Sm. 

geswdgen ptc swoimen vb ewoon tswuwn 

' ienaelen.' aswOgen ' choked * — swongh sbit, iswowen ' in a swoon ' — 
(sann) Ld, (sunn) Beh, Sh, 

wdgian wowen woo wuw 

(nu) Jn. 

genOg tinoh enough inaf 

— pi inojhe 0, inouh AB, inog Ay, inoch, enogh CM, enewe : knew, 
enoghe : soghe (^ sugu) TM — ^ynough Td. (innz, inuf) G, (invf) W, 
(en«f) Cp. pi enow (enaa) W, Cp, (envf) Beh, (fisawt) 8h. pi (enin) 
Bihy (eenoa) iS%. 

2120 dr5g prt tdroh(h) drew druw 

— drotth AB, drogh North, drow : prow ' profit ' BBC, dro$, droa(5), 
dreu5 Wiel,—4iu9 Td. 

pldh tploh plough plau 

from Scand. plOg — ploges Ld, plo:do TM. plow (plouh) PBL 
plough : ynongh, plowman C9i— (eu) Cp, (00) Jn, (on) Beh, Sh. 

bdg tbojheaa pi bough bau 

— bowes : growes, bughea PC — (bowh, buuh) BU, (boa) O, (beu, boo) 
Jn. (boo) Beh. (boa) Sh, 

rot tr6te root ruwt 

from Seand. r9t (?) unroUioe dob ' exterminant ' Wgl — rotfest Ld, role 
AB, 

adt sot Boot 8ut, sat 

(uu) O, (uu) Pr, (u) Cp. (•) heUer (a) Jn. (•) Beh, Sh, 

(svU) Beh, (suati) Sh soofy. 

2125 wTotaa wroten root ruwt 

' root up.' wrOi ' snoot.' 

fot tf6t foot fiit 
(uu) BU, Pr, (u, «) Jn. (•) Beh. (u) Sh. 

gem5t sn mot moot muwt 

' meeting.' 

bot sf tbote boot buwt 

( mending.' 

had hod hood hud 

(u, uu) 8m, (u, •) Jn. 

. - * j_ - (rood ruwd 

2130 rod sf trode < _ _ 

( rod rod 

^lodde : oodde ON-^rodd 'rod* Td. 
geeodd ptc iaohood shod Jod 

6 b 2 



372 



HISTOEY OP ENGLISH SOUNDS. 



Bt5d prt tfltod Btood 
(uu) G. (u, •) Jn, (uu) Bek. (u) Sk. 

wddnes-dflBg wednesdai Wednesday 

— wodnesdawes (weodneidei) AR, weduesdai Kath, 
— (wenzdee) /», 8h. (wensdi) Bch. 

f5da tfode food 

2135 f5d(d)or fod(d)er fodder 

Add tflod flood 

flttdda Td, (uu) 8m, (u) BU, Q, («, •) Cp. 

indd imod mood 

twer-mdd wermod wormwood 

mddor tmoderr mother 

— ^moder North,, AucL mooder Ch — ^mwddyr HVg, 

2140 g6d tgod good 

— guod Ay, goud AllP, — gwd EVff. (uu, u) Sm. 
brod brod brood 

blod tblod blood 
bloud Td, Ch, (uu) Sm, (u) Bll, Q, 

hwdpan — whoop 
(huup, uup) Jn, 



stud 

we(d)xiBdi 
Wednesday HarU 

fiiwd 
foder 
flad 

muwd 
weemwud 

maCer 

00 Ck. (u) BU. 

gud 

(u) G. (u, «) Jn. 
bruwd 
blad 

(w)huwp 



SECOND WORD-LIST. 







(LIVINO-OLD.) 








a 




av 

:avi 


shore 


scQfan 








shovel 


soofel 


JMure 
wari 


thorough 

worry 


)mrb 
wyrgan 


fflav 
foksglav 


glove 
foxglove 


dof 

foxes-glofa 


faron 


furrow 


farh 


obaT 


above 


onbufan 


baro 


borough 


burg 














hai) 


hung 


hSng 


d$X 


dull 


dol 


j»9 


young 


(hrung 
) wrungen 


halk 


huUk 


hulo 


»9 


rung 


najni) 


nothing 


nanlmig 


Ui, 


lung 


lungen 


da> 


doth 


d&> 


sai) 


eung 


sangen 




^ 


slai) 


slung 


slungin 


a0ar 


other 


0])er 


swai) 


swung 


swungen 


satgn 
rnavw 


southern 


sQ^eme 


stai) 


stung 


stungen 


mother 


mS)>or 


sprai) 


sprung 


sprungen 


braOar 


brother 


br0)K>r 


eman(8t) 


among 


ongemang 








kUi) 


elung 


olungen 


as 


ue 


as 


tax) 


tongue 


tunge 


8aB 


thus 


>a8 


dag 


dung 


dung 


masl 


mused 


muscle 


sank 


sunk 


sunoen 


basl 


buede 


bustla 


fpai)k 


shrunk 


scruncen 


task 


tusk 


tuso 


ma^k 


monk 


munuc 


bask 


busk 


bOask 


dra^k(oii' 


) drunk{en) 


drunoen 


bastii)! 


huelings 


has^g 


baqgar 


hunger 


bnnger 


rast 


ruet 


nut 


maijger 


monger 


mangere 


last 


lust 


lust 








Inrast 


thrust 


>ry8ta 
( must 
I mOste 


hani 


honey 


hunig 
( eoman 
( umen 


mast 


mitsi 


ran 


run 


klastor 


cluster 


clyster 


san 


eon 


sunu 


gast 


gust 


gust 


san 


sun 


•unne 


trast 


trust 


treysta 


Stan 


stun 


stunian 


dast 


duH 


dOst 


span 


spun 


spunnen 








Jan 


shun 


Bounian 


hazi(f) 


hussy, -if 


haswif 


wan 


won 


gewunnen 


hazbond 


husband 


hQsbOnda 


wan 


one 


ftn 








nan 


nun 


nunne 


»f 


rush 


rinc 


nan 


none 


nftn 


bla 


thrush 


yryeoe 


kanii) 


cunning 


onnnan 


blu^ 


blyscan 


bigan 


begun 


begunnen 








tan 


tun 


tunne 


raf 


rough 


rab 


dan 


dun 


dnnn 


wndraf 


woodruff 


wudurGfe 


dan 


done 


gedOn 


inaf 


enough 


genSg 


man]) 


month 


mdna^ 


kaf 


cuff 


cuffie 


wans 


once 


vne 


Uf 


tough 


t5h 


bant 


hunt 


huntian 




■^ 




stant 


stuni 


stunt 


avn 


oven 


ofen 


pant 


punt 


punt 


lav 


lore 


lufian 


andor 


under 


under 



! 



374 



HISTORY OF ENGLISH SOUNDS. 



handred 

)»and9r 

sandi 

wandar 

inandi 

irandl 

bandl 

]»am 

Bam 

samar 

swam 

nam 

kam 

kamli 

kram 

gam 

dam 

plam 

krampl 

slambdr 

tambl 

Bak 
liiMiitmk l 

mak 
klak 
plak 
bak 

agli 

magwaot 

tag 

ator 

fat 

Jatl 

flatar 

nat 

katlfij 

gat 

bat 

bat 

batar 

batak 

BatJ 

matj 

kratj 

adar 

radar 

radi 

flad 

kad 

blad 

»P 
Bap 

kap 

tapanB 

Btabl 



hundred 

thunder 

eunday 

wonder 

monday 

trundle 

"bundle 

thumb 

eome 

tummer 

9wum 

numb 

come 

comely 

crumb 



hundred 

)>unor 

Bunnandffig 

wundor 

mSnandsBg 

tryndel 

byndelle 

^ma 

Bum 

Bumor 

swummen 

genumen 

ouman 

cjmlic 

cruma 



gum 
dumb 


g5ma 
dumb 


plum 

crumple 

slumber 


plame 
crump 
Bluma 


tumble 


tumbian 


suck 


BQcan 


honeysuckle 

muck 

duck 


hunigBOge 

myk 

cloodan 


pluck 
(mck 


pluccian 
buoca 



ugly 

muffwort 

tug 

utter 
shut 
shuttle 
flutter 
iiut 

cuttlefish 
gut 
but 
butt 
butter 
buttock 
such 
much 
crutch 

udder 
rudder 
ruddg 
flood 
cud 
blood 

up 
sup 
cup 
twopence 

stubble 



ugglig 

mucgwyrt 

togian 

aterre 

Bcyttan 

Bcytel 

floterian 

hnutu 

cudele 

gutt 

batan 

potian 

butere 

buttuc 

Bwelo 

micel 

cry 00 

tider 

rojjor 

rudig 

fl5d 

cwidu 

bl5d 

tip 

' BQpan 
cuppe 
twap^ningas 

Btybb 



e 



a(n) 

-dam 

-ed 
-wad 



8mi)> 

ki]> 

pil) 

hi^ar 
»i«ar 
Bmit(i 
Bti^i 



a{n) 

-dom 

'herd 
-ward 



Btirap stirrup 



il 


m 


hU 


hitt 


])ril 


thrill 


Bil 


sOl 


sili 


silly 


Bwil 


swiU 


Bkil 


skill 


Btil 


stia 


jTlm 


SfdU 
shilling 


wil 


wiU 


wilou 


willow 


fil 


fill 


mil 


mat 


kil 


kill 


tu 


till 


tfU 


ehiU 


dil 


diU 


pilou 


pillow 


bil 


bill 


bilou 


billow 


fil> 


fiUh 


milf 


mSLch 


silvar 


siher 


kiln 


kUn 


film 


film 


Bilk 


silk 


milk 


milk 


hilt 


hilt 


gilt 


guilt 


mildjuw 


mildew 


tjildren 


children 


gild 


guild 


gild 


gild 


bUd 


Jmild 



smith 

kith 

pith 

hither 
thither 
smithy 
stithy 



fin 

-dom 

-hirde 
-weard 



■tigrftp 

ill 

hyll 

)>yrelian 

ayll 

ge8£lig(S) 

Bwilian 

Bkil 

Btille 

spildan 

Bcilling 

willa 

welig 

fyllan 

mylen 

owellan 
\tU 
] tilian 

Ofle 

dile 

pyle 
}bUe 
(Inll 

bylgja 

m 

milce 

seolfor 

cylen 

filmen 

Beoloc 

meolo 

hilt 

gyit 

mildSaw 

cdldru 

gildi 

gyldan 

byldan 

mnip 

cyJ)l)o 

pi)>a 

hider 
)nder 
Bmi])]ie 
Bt^ln 



SECOND WOED-LIST. 



375 



WltJl 

whiter 

-Us 

lign 

Jrisl 

Oil 

whifll 

mis 

mislioa 

kii 



kriindam 

krifmoi 

grid 

glisn 

brial 

bill 

risi 

lilt 

litUii 

sitter 

fist 

mist 

grist 

twist 

distaaf 

lisp 

whisper 

Jorisp 

• 

IS 

lus 

rizn 

wizn 

ffrizli 

oizi 

bid 

wisdom 

? 

dij 
bijap 

if 

stif 

klif 

fif> 

lift 

)nift 

sift 

swift 

fift 

frift 

fifti 

Rift 
drift 



with 

ftithy 

whither 

"lets 

Usten 

thistle 

this 

fohistle 

misM 

mistletoe 

hin 

chriHen 

ehriHendom 

ehfHstnuu 

griHle 

glisten 

bristle 

bliss 

wrist 

list 

listless 
sister 

jut 

mist 

grist 

twist 

distaff 

lisp 

whisper 

crisp 

• 

u 

his 

risef^ 

wizen 

grisUf 

ditty 

busy 

wisdom 

wish 
fish 
dish 
bishop 

if 
•tiff 

diff 
j^fth 

drift 

sift 

swift 

shift 

shrift 

fifty 
ffift 

drift 



wijriff 
hwider 

-ISm 

hlysnui 

Jristel 

]ris 

wistliAn 



misteltSn 
oyssftn 
eristenian 
oristendSm 



gristle 

ffljmiAii 

byrst 

bliss 

wrist 
(ffelystan 
( nlystan 

lust 

sweostor 

ffst 

mist 

grist 

twist 

distef 

wlisp 

hwisprian 

crisp 

is 

his 

risen 

wisnisn 

ongrislio 

dyng 

bysig 

irisd5m 



fiso 

disc 

biscop 

gif 

stif 

dif 

nfta 

lyfta 

Jnrift 

siftan 

swift 

skifta 

scrift 

riftig 

Kift 
drift 



liT 

liver 
siv 
giv 
driTn 



rii} 

sig 

slix) 

swii) 

•^ 

Btni) 

sprii) 

win 

kip 

kill) 

brii) 

rinkl 

lii)k 

)rfi)k 

siqk 

slii)k 

stix)k 

Jrjgk 

wii)k 

twix)kl 

drii)k 

inglend 

finger 

mii)gl 

in 

in 

linin 

Unit 

]iin 

sin 

sinja 

skin 

spin 

win 

winon 

fin 

minoQ 

kin 

bigin 

grm 

tin 

tpn 

twin 

din 

bin 

bin 

linsijd 

sins 



live 

liver 

sieve 

give 

driven 



r%ng 

wring 

thing 

»ng 

sling 

swing 

sting 

string 

spring 

wing 

king 

ding 

bring 

wrinkle 

link 

think 

sink 

slink 

stink 

shrink 

wink 

twinkle 

drink 

english 

england 

finger 

mingle 

in 

in, inm 

linen 

linnet 

thin 

sin 

sinew 

skin 

spin 

shin 

win 

winnow 

fin 

minnow 

kin 

begin 

grin 

tin 

chin 

twin 

din 

bin 

been 

lin$eed 

since 



libban 

lifer 

sife 

Sibil 

dnfen 

t bring 
(h)ringui 
wringan 
Jnng 
singan 
slongya 
Bwingan 
stingan 
strange 
springan 

<nFning 

dingan 

bringan 

wrinde 

hl^noe 

^ynoan 

sincan 

slinoan 

stinoan 

sorincan 

winoian 

twinolian 

drincan 

fnglisc 

^ngla«laad 

finger 

m^ngan 

in 

inn 

linen 

linetwige 

jTynne 

sj^nn 

Sinn 

ddnn 

spinnan 

sdna 

gewinnan 

windwian 

finn 

myne 

cynn 

beginnan 

grennian 

tin 

cinn 

Setwinn 
yne 
binn 
*gebe<m 
Unsffd 



76 



HISTORY OP ENGLISH SOUNDS. 



minstar 

linjpin 

wIdJ 

finj 

nuS 

kriii5 

stint 

wintar 

flint 

mint 

dint 

hinddr 

lindin 

sinddr 

Bpindl 

wind 

windou 

windlas 

tindar 

dwindl 

him 

rim 

lim 

swim 

Jimer 

wimin 

grim 

trim 

dim 

brim 

brimstan 

imp 

]imp 

)>imbl 

nimbi 

timbar 

(aia)ikl 

rik 

lik 

(gaa)lik 

>ik 

sik 

dkl 

stik 

striken 

wik 

wikid 

fikl 

flikar 

kwik 

kwikseend 

tik 

tiki 

tjikin 

prik 



mineter 
inch 

UnchCptn) 
winch 
finch 
tinge 
cringe 
ftint 
winter 
flint 

mint 

dint 

hinder 

linden 

cinder 

spindle 

wind 

window 

windlass 

tinder 

dwindle 

him 

rim 

limb 

swim 

shimmer 

women 

grim 

trim 

dim 

brim 

brimstone 

imp 

limp 

thimble 

nimble 

timbir 

icicle 

rich 

lick 

garlic 

ihick 

sick 

sickle 

stick 

stricken 
wick 
wicked 
flckle 
flicker 
quick 
quicksand 
tick 
tickle 
chicken 
prick 



mynster 

ynce 

lynes 

wince 

fino 

B^ngan 

cringan 

styntan 

winter 

flint 

minte 

mynet 

dynt 

hindrian 

linden 

Binder 

spinel 

wind 

Tind9aga 

vindia 

tynder 

dwinan 

him 

rima 

lim 

iwimman 

scimerian 

wifm^nn 

gnmm 

trymman 

dimm 

brymme 

*brynestftn 

impa 

l^p(healt) 

Jjymel 

numol 

timber 

gecel 

hrSac 

liocian 

(g5r)lgao 

J>yooe 

sSoo 

sicol 
I Btician 
[ sticca 

stricen 

weooe 

wioce 

ficol 

flicorian 

cwic 

cw^cesand 

ticia 

citelian 

dcen 

prician 



prikl 

siks 

viksn 

miks 

siks)) 

bitwikst 

iawig 
twig 

it 

hit 

ritn 

litl 

sit 

slit 

smitn 

spit 

wit 

whit 

fit 

flit 

nit 

grit 

glitar 

twit 

pit 

priti 

bit 

bitn 

bitar 

itj 

BtitJ 

witj(elm) 

witj 

whitj 

flitf 

kitjin 

twitf 

ditj 

pitf 

bitj 

britjiz 

hid 

rid 

ridl 

ridn 

(bed)ridn 

lid 

slid 

widon 

fidl 

midl 

kid 

kwid 

gidi 

did 



prickle 

six 

vixen 

mix 

sixth 

bettaixt 

earwig 
twig 

it 

hU 

written 

little 

sit 

slit 

smitten 

spit 
wit 



prioel 

sex 

fyxen 

miecian 

sexta 

betwix 

Sanxricga 
twig 

hit 

hitta 

writen 

lytel 

sittan 

sliten 

smiten 
( spita 
( spittan 

Iwitan 
gewitt 



whit 


wiht 


flt 


fltt 


flit 


flytja 


knit 


cnyttan 


grit 


grytt 


glitter 


glitter 


twit 


setwitan 


pit 


pytt 


pretty 


prsettig 


bit 


bite 


bitten 


biten 


bitter 


biter 


itch 


giccan 


stitch 


stioe 


wychelm 


wice 


witch 


¥ricce 


which 


hwilo 


flitch 


flicce 


kitchen 


C3rcene 


twitch 


twiocian 


ditch 


die 


pitch 


pic 


bitch 


bicce 


breeches 


brcec 


hid 


hydde 


rid 


hr^dan 


riddle 


reedels (6) 


ridden 


riden 


bedridden 


b^ddrida 


lid 


hlid 


slid 


sliden 


widow 


widwe 


flddU 


fi])ele 


middle 


middel 


kid 


ki> 


quid 


cwidn 


giddy 


gidig 


did 


dyde 



SECOND WOED-LIST. 



377 



bid 


^v; 


biddan - 


hel)> 


health 


h^lo 


bidn 


bidden 


biden 


ehi 


elee 


^les 


midst 


midtt 


tOmiddes 


welf 


weleh 


weliso 


rid3 


ridge 


hrycg 


elf 


elf 


«lf 


ndds 


midf/e 


mycg 


Belf 


eelf 


self 


bridg 


bridge 


brycg 


Jelf 


%hdf 
twelfth 


BO^lf 


i_ • 


« p 


A 


twelf> 


tw^lfla 


hip 


hxp 


hype 


twelv 


twelve 


twelf 


hi|M 


hipe 


hSopan 


delv 


delve 


delfian 


"? 


lip 


lippa 


elm 


elm 


elm 


^npens 


threepenet 
eowMiip 


>r6op^ngM 


helm 


helm 


helma 


(^kau^slip 


cflslyppe 


helmit 


helmet 


helm 


■lipori 


tUppery 


fllipor 


welkin 


welkin 


wolcen 


■trip 


etrip 


BirepAn 


whelk 


whelk 


weoloo 


H , 


ship 


8oip 


smelt 


emelt 


emelt 


kripl 


cripple 


.S?^^ 


Bpelt 


tpelt 


Bpelt 


klip 


clip 


{ klippa 
I olyppan 


felt 


feU 


felt 


m 


F 


melt 


melt 


meltan 


tipit 

• 


tippet 


taeppet 


belt 


belt 


b^lt 


grip 


ffrip 


gripe 


beltf 


heleh^ 


beloettan 


dip 


dip 


dyppan 


eldar, -ist 


elder, -est 


fldra, ^dest 


drip 


drip 


drypan 


eldar 


elder 


ellem 


rib 
nib 


rib 
nib 


ribb 
n^bb 


held 

Beldam 

help 


held 

seldom 

help 


hSold 

seldon 

helpan 




gk 




jelp 


yelp 


gelpan 
hw^lp 




V 




whelp 


yfhelp 


erend 




vrende (9) 








herin 


herring 


hsring 


de> 


death 


d6a> 


Jerif 


sheriff 


■drgerogfa 


bre> 


breath 


brft> (6) 


fen 


ferry 


ffrian 








meri 


merry 


myig 


IcCar 


leather 


le]>er 


beri 


oerry 
bury 


b^rge 

byrgan 

byreels 


wetter 


wether 


we>er 


beri 


we'Sar 


weather 


weder 


berial 


burial 


wheVar 


whether 


hws^ 






•r O 


fetter 


feather 


fe)wr 


el 


ell 


«ln 


neOer 


nether 


Beo))or 


hel 


heU 


hfll 


tageiSer 


together 


tOgaedre 


jel 


yea 


gellan 


bieGrin 


brethren 


broe^ 


jeloQ 


yellow 


geolu 








Ml 


eell 


B^an 


jee 


yes 


giBe 


■wel 


swell 


Bwellan 


reel 


wrestle 


wrSstban 


Broel 


emell 


Bmellan 


les 


less 


IfiBM 


r 


epM 


Bpell 


kres 


cress 


creese 


eheU 


BC^U 


bles 


hUss 


blSdflian 


1 


ioell 


(wel 
i w«lle 


bihest 


behest 


hfts 


wel 


jeitedi 


yesterday 


geostran-dseg 






(fell 


rest 


rest 


r^st 


fel 


fea 


\ ffllan 


rest 


wresi 


wrSetan 






(f^U 


lert 


lest 


>7-l*s 


fell, -on 


felly, 'oe 


f«lg 


weet 


west 


west 


felou 


fellow 


ftli«i 


nett 


nest 


nest 


nel 


knea 


onyllan 


geBt 


guest 


e^ 


kwel 


qwfl 


cwfUan 


tjeat 


ehesi 


oest 


tel 


teU 


tfllan 


beet 


best 


bftst 


dwel 


dwell 


dv^lja 


breBt 


breast 


brtoet 


bel 


beU 


belle 








bell 
belotiz 


btUy 
bellowt 


b?lg 


sez 


says 


■««•> 


belott 


bellow 


belgan 


{vejotild 


. threshold 


)>erBcold 



378 



HISTORY OF ENGLISH SOUNDS. 



fre 
fie 



1 



hefor 

def 

bireft 

left 

>eft 
weft 
kleft 
deft 

eTer 

evii 

heTi 

hevn 

ilevn 

■eyn 

nevor 

devl 

lengp 
8trei))» 

eni 

hen 

renit 

Ten 

iSen 

wen 

when 

fen 

men 

meni 

ken 

eg6n(8t) 

ten 

den 

pen 

pen! 

hens 

Vena 

whena 

klenz 

fitenf 

renj 

frenj 

kwen 

dren 

ben 

enent 

lent 

lent 

sent 

ment 

kent 

twenti 

end 

rend 



enf 

f 



freA 
Jle$k 

heifer 

deaf 

hereft 

left 

ihefi 
v^eft 
cleft 
deft 

ever 

evet^ 

heavy 

heaven 

eleven 

eeven 

never 

devil 

length 
strenffth 

any 
hen 
rennet 
wren 
then 
wen 
when 
fen 
f}i0fi 
many 
hen 

againei 
ten 
den 
pen 
penny 
hence 
thence 
whence 
cleanse 
stench 
wrench 
french 
quench 
drench 
bench 
anent 
lent 
lent 
sent 
meant 
kent 
twenty 
end 
rend 



fenc 
fl«w 

hSahfore 

demf 

bereafbd 

(l«fed 

wefta 

gedyfte 

gedsfte 

£fre 

fifire tio 

h^fiff 

heofon 

endlufon 

■eofon 

nsfre 

ddofol 

l^ngo 
s^ngjN) 

etnig 

h^nn 

r^nnan 

wrfinna 

)K)nne 

w^nn 

hwonne 

fignn 

m^nn 

manig 

k^nna 

ongsBgn 

tSn 

d^nn 

p^nnan 

pining 

heonon 

J^anon 

hwanon 

clAnsian 

it^nc 

wr^can 

fr^nciflc 

cw^ncan 

dr^ncan 

b^nc 

on efen 

l^oten 

Isned 

tend 

mftned 

c^t 

twentig 

^de 

randan 



lend 


lend 


Isnan 


aend 


send 


s^ndan 


spend 


spend 


sp^ndaji 


wend 


wend 


w^ndan 


frend 


friend 


frtend 


bend 


bend 


bfndan 


blend 


blend 


blfndan 


hem 


hem 


h^mm 


hemlok 


hemlock 


hymlic 


lemen 


lemman 


liof maim 


Oem 


them 


)»eim 


stem 


stem 


stfmma 


stem 


stem 


st^mn 


emti 


empty 


Amettig 


hemp 


hemp 


hnnep 


embes 


embers 


eimyrja 


rek 


reek 


r^ocan 


rek 


wreck 


VT^ 


rek 


wreak 


wreoan 


rekan 


reckon 


receniaa 


spek 


spedc 


speooa 


nek 


neck 


hn^oea 


bek(en) 


beel(on) 


bS^ian 


nekst 


next 


nShst 


f» 


^S 


«g 


l*« 


1^9 


1«« 


dregs 


dregs 


di^ 


beg 


beg 


bededan 



et 
jet 

let 

]xret(n) 

set 

setl 

swet 

wet 

whet 

fetar 

fret 

net 

netl 

met 

keU 

get 

teter 

betar 

retj 

retj 

stretf 

fetj 

hed 
red 
red 



ate 
yet 

let 

threai{en) 
set 

seHle 
sweat 
wet 
whet 
fetter 

fret 

net 

nettle 

met 

kettU 

yet 

tetter 

better 

retch 

wretch 

stretch 

feieh 

head 

red 

read 



«t 

get 

! rattan 
lAtan 
let 

)fr6atiaii 

B^taa 

■eU 

swfttan 

wAt 

hw^tan 

fetor 
Ifrtan 
( fruftwan 

nftt 

n^tele 

gemStte 

c«tel 

(be)getan 

teter 

b^iera 



WT^ooa 

■tr^oean 
f((ooan 

heafod 

read 

rftdde 



SECOND WOED-LIST. 



379 



led 


lead 


UtA 


kelon 


eoilow 


oalu 


led 


led 


lAdde 


g»l0U2 


gallows 


g»lg» 


]ved 


thread 


]n«d(6) 


taelou 


tallow 


tKlff 

salfian 


Bed 


said 


ssBgde 


Sffilv 


salve 


Bted 


Head 


st^de 








Btedi 


steady 


StCB)>^ 


b»> 


halh 


hiBf> 


sped 


sped 


spo&dde 








spied 


spread 


Bpradan 
sprSdde 


IseSer 

fSdVBTR 


lather 
fathom 


lea>or 
heyiR 


Jed 


shed 


Bcydd 
BoSadan 


geSar 


gather 


gsdrian 


Jred 


shred 


screadiaa 


mss 


mass 


msBse 


wed 


ved 


wedd 


b«s 


bass 


bflBrs 


wednzdi 


Wednesday 


wOdnes-dng 


bSBSt 


bast 


bSMt 


fed 


fed 


foedde 


nspin 


aspen 


nspe 


medon 


meadow 


in«d(9) 








trod 


tred 


tredan 


IBS 


as 


alBwS 


ded 


dead 


dead 


hes 


Aa# 


htBf> 


died 


dread 


ondrftdaa (9) 








bed 


bed 


b^d 


^ 

» 


ash 


CBBC 


bred 


bread 


bread 


ejiz 


ashes 


ascan 


bled 


bled 


blcedde 


rsej 


rash 


nsk 


•dj 


edge 


W 


fDsJ 


thrash 


>er8ean 


hed^ 


hedge 


h«e 


mash 


nUBBO 


•^^ 


sedge 


HCg 








8led2(h»m-«^^tf 


■Ifcg 


tjsfer 


chaffer 


*ceap&ra 


wedg 


wedge 


w«cg 


bSBT 


have 


habban 


Btep 


step 


stfppan 


htei) 


hang 


hikfigfiyfi 


rtep- 


stqi>' 


steop- 


SSI) 


sang 


sang 


Jepod 


shepherd 


BcApbiide (6) 


sprsBi) 


sprang 


sprang 


wepen 


weapon 


wftpen (e) 


g»9 


gang 


gang 


dep> 


depth 


depe 


bei) 


bang 


banga 


slept 


slept 


Bl«pte(e) 


sx)ker 


anchor 


anoor 


kept 


kept 


OOBDte 


ei)kl 


ancle 


anoleow 


m 


A 


a 


naii)k 


hank 


hanki 


eb 


ebb 


«bU 


r»i}k 


rank 


rano 


web 


w^ 


w^bb 


Iftgk 


lank 


hlano 


pebl 


pebble 


papol 


)«i)k 


thank 


Randan 








WSB^k. 


sank 


Banc 








Btenk 
fogk 


stank 


Btanc 




m 




shank 


Bcanca 








Tnegk 


shrank 


Bcrano 


mrovL 


arrow 


arwe 


knegk 


crank 


orano 


hmri 


harry 


h^rgian 


dneijk 


drank 


drano 


jasroa 


yarrow 


gearwe 


b«i)k 


bank 


banki 


Bperon 


sparrow 


spearwa 


si)g9r 


anger 


ADgr 


nsroa 


narrow 


neam 


«a|gl 


angle 


angel 


nueroQ 


marrow 


mearg 








tnri 


tarry 


MtS^ 


mn 


ran 


am 


bsBroa 


barrow 


Jbearwe 
jbeoig 


Bpcen 


than 
span 


)K>nne 
i spann 
( Bpannan 


lueloa 


hallow 


hSlgian 


fsm 


fan 


(knn 


Bselon 
J»l 


sallow 
$hal 


Bain 
Balh. 

BC»1 


nuen 
ksn 


man 
can 


mann 

cann 

oanne 


lieeloQ 


fallow 


fala 


bigsen 


began 


b^gann 


m»lou 


mallow 


malwe 


ksanl 


eannel 


eandel 



38o 



HISTORY OP ENGLISH SOUNDS. 



gsenit 
ten 
psen 
bsen 

nensek 

snvil 

end 

hsend 

laend 

■send 

Btsnd 

Rtrsend 

ksendl 

gsendgr 

bnend 

fern 

hem 

haamer 

rsem 

Item 

Isemaa 

Bwsem 

stemar 

knem 

Fteemp 

krsemp 

biiembl 

Isk 

88ek 

nenssek 

slsek 

J»kl 

knek 

bek 

blfek 

seka 

seksl 

wffks 

flseks 

r»g 



wteg 
baeg 

»t 

hst 

Iteter 

»fet 

Bset 

Fietedi 

spet 

fset 

flaet 

Viet 

met 



gaitnet 

fan 

pan 

bann 

anthetn 

rantack 

anvil 

and 

hand 

land 

sand 

stand 

strand 

candle 

gander 

brand 

am 

ham 

hammer 

ram 

lanib 

lammas 

svam 

slammer 

cram 

stamp 

cramp 

bramble 

lach 

sack 

ransack 

sUick 

shackle 

crack 

back 

black 

axe 

axle 

vox 

fiax 

rag 
shag 
wag 
bag 

at 

hU 

latter 

that 

sat 

Saturday 

spat 

fat 

flat 

rat 

gnat 



ganot 

tannian 

panne 

gebann 

antefii 

rannsaka 

aniUt 

and 

hand 

land 

Band 

standan 

strand 

candel 

gandra 

brand 

eom 

hamm 

hamor 

ramm 

lamb 

blftfnuesBe 

Bwanun 

Btamrian 

cranunian 

Btampian 

cramp 

bngmel 

lak 

BSBOC 

rannsaka 
flee 
scacol 
cracian 

bSBC 
bllBO 

sx 

fexl 

waexan 

W8PX 

fliex 

raggig 
sceacga 
wagian 
ba^ 

8Pt 

hcett 

lator 

]«t 

Met 

sKtemeS'dieg 

spe&tte 

f^t 

flat 

faet 

gnaett 



mstok 

kset 

klsetar 

bigiet 

tietdr 

betn 

laetj 

J»tf 

msBtJ 

eddr 

sedld 

hsd 

liedor 

seed 

BSBdl 

Jaedou 

msed 

mseddr 

g8ed(flai) 

klsed 

glasd 

bsBd 

biBd 

bliedar 

aedz 

sepl 

hcep 

Isp 

l«p, -it 

Ispwii) 

raep 

Btrtep 

nsep 

ksep 

klsep 

t«p 

trep 

tjsepman 

ebat 

Bksab 

Jflebi 

kr»b 

gKb(l) 



tu 

wul 

ful 

fular 

pul 

bal 

bulak 

wulf 

bnsam 



mattock 


mattoo 


cat 


oatte 


< latter 


datrian 


begat 


begnt 


tatter 


ta»ttec- 


batten 


batna 


latch 


gelseocan 


thatch 


Jneo 


match 


gem^oca 


adder 


n£dre (e) 


addled 


adela 


had 


luefde 


ladder 


hlfidre 


sad 


Bsd 


saddle 


sadol 


shadow 


Bceadn 


mad 


gemftdd 


madder 


nuedere 


SlodiM 


gadd 


dad 


klfijKU 


glad 


glaed 


bade 


bed 


bad 


bfiddel 


bladder 


bladre(S) 


adze 


adese 


apple 


»ppel 


hap 


happ 


lap 


lapian 


lap, -pet 


lappa 


lapwing 


Inpewince 


sap 


8«p 


strap 


Btropp 


nap 


hnappian 


cap 


cseppe 


dap 


klappa 


tap 


teppe 


trap 


traeppe 


chapman 


cSapmann 


abbot 


abbod 


scab ) 
shabby ) 


■oebb 


crab 


crabba 


gab{ble) 


gabba 



to 



td 



tfool 


wall 


full 
fuller 

puU 

bull 


full 
fullere 
puUian 
buli 


buUodt 


bulluc 


wolf 


wulf 



bosom 



^sm 



SECOND WOBD-LIST. 



381 



huf 


koof 


hOf 


wotp 


wasp 


wasp 


_ . / \ 






gospl 


gospel 


godfpell 


■pu(w)ii 


tpoon 


spOn 








mm 


room 


rOm 


WOI 


was 


W»8 


wamon 


wman 


wlfnuum 


woj 


m 




bra(w)m 


hroam 


brOm 


wa$h 


w»8oan 


hak 


hoOi 


hOo 


ofl 


offal 


offall 


ruk 


rook 


hrOo 




^ 


^ 


luk 


look 


lOoUn 


ov 


of 


of 


fuk 
knk 


AJkXMb 


80O0 


hOTl 


hovel 


^hofel 


cook 


oOc 


grovl 


grovel 


grQfa 


kruk 


orook 


kiok 


proToit 


provost 


prftfost 


ink 


took 


too 








bok 


book 


boo 


roi| 


wrong 


vrang 






jbrOo 


loi) 


long 


lang 


bruk 


brook 


J)0I) 


thong 


>wang 






Jwroij 


throng 


ge)nrang 


■at 


toot 


BOt 


■01) 


song 


Mkng 


fat 


foot 


fbt 


■troi) 


etrong 


Strang 






t0I)S 


Umg$ 


tang 


hud 


hood 


hod 








-hod 


•hood 


-had 


on 


on 


on 


stad 


$tood 


BtOd 


opon 


upon 


nppon 


Jad 
wad 


should 


Boolde 


anon 

• 


anon 


onftn 


tpood 


wadu 


jon 


yon 


geon 


wad 


toould 


wolde 


■won 


swan 


swan 


kud 


eould 


oQ>e 


Jon 


ehone 


Bc&n 


gad 


good 


gOd 


won 
gon 


wan 
gone 


wann 
g«g*n 








wont 


wont 


vanta 









wonton 


wanton 


^wantogen 


• 




^ » 


bijond 


beyond 


begeondan 


son 


$orrg 


flftrig 


wond 


wand 


V9nd 


aoroa 


sorrow 


■org 


wonder 


wander 


wandrian 


moroa 


morrow 


morgon 


bond 


bond 


band 


boroa 


borrow 


boi^^Aii 














from 


from 


from 


hoU 


hoUy 


hol^gn 




■f 




hoUdi 


holiday 


hfiligdng 


(hoH)hok 


{hoUy)hoek 


hooc 


holoa 


hollow 


holh 


hok 


hough 


hOh 


swoloa 
woloa 


swallow 
wallow 


( swalwe 
( swelgan 
walwUn 


rok 
lok 


rock 
lock 


rooo 
Hoc 
(looc 


foloa 


follow 


folgiui 


■ok 


BOCk 


BOOO 


nolidj 


knowUdgs 


'cnftwlAcan 


■mok 


emoek 


smoc 


olBoa 


also 


AltWft 


■tok 


stoek 


stooo 


foU 


false 


UX% 


flok 


flock 


floco 


holt 


haU 


halt 


nok 


knock 


cnocian 


holtor 


haUer 


halfter 


kok 


cock 


0000 


■olt 


$aU 


uJt 


kokl 


cockle 


oocoel 


molt 


malt 


malt 


kiokdri 


crockery 


crooca 








dok 


dock 


dooce 


mos 


mots 


mon 


oks 


oxe 


oxa 


gonp 


go$sip 


godsibb 


foks 


f^ 


fox 


gothak 
drot 


goshawk 
dross 


gOahftfoc 
drome 


poks 
boka 


C 


poooaa 

box 


bloaam 


hiossom 


blOBtme 








foet»r 


fodsr 


fttetor 


frog 


frog 


frogga 


noitril 


nostril 


luet^rel 


dog 


dog 


duoga 



382 



HfSTOBY OF ENGLISH SOUNDS. 



otar 

hoi 

rot 

lot 

■not 

■pot 

Jot 

wot 

wotl 

whot 

not 

not 

kot 

klot 

£?*♦ 
bigotn 

dot 

plot 

botom 

blot 

wotj 

od 

rod 

■odn 

swodl 

Jod 

fodor 

kod 

god 

tiodn 

bodi 

hop 

■op 

■top 

strop 

kopor 

krop 

top 

drop 

popi 

kob(web) 
lobster 



oiler 

hot 

rot 

lot 

inoi 

$pot 

thiol 

wot 

teaitlo 

tehai 

not 

hnot 

eoi 

dot 

got 

begotten 

dot 

phi 

bottom 

hlot 

teatch 

odd 

rod 

eodden 

ewaddle 

ehod 

fodder 

eod 

god 

trodden 

body 

hop 

top 

etop 

strop 

copper 

crop 

top 

drop 

poppy 

e6b{web) 
lobeter 



a«r 

StMT 

spaar 
faar 

taftr 

tjaar 

9d5 



are 
star 
spar 
far 
mar 
tar 
char I 
ajar \ 



otor 

hSt 

rotian 

hlot 

gosnot 

splott 

(gOKSOt 

) aooten 
w&t 
watdl 
hwaet 
nfiwiht 
cnotta 
cot 
clott 
(be)gKt 
begeten 
doU 
plot 
botm 
plot 



odd! 

red 

■oden 

8W9]>el 

gesoQd 

fisdor 

codd 

god 

troden 

bodig 

hoppian 

Boppian 

stoppian 

■trop 

oopor 

icropp 
kroppa 
topp 
dropa 
popig 

(fttor)coppa 
loppestre 



earon 

■taorra 

Bpeer(BtSn) 

feorr 

m^rran 

teom 



gaalik 
tfaalak 
daalii) 
baaU 

haa> 
laaj^ 
paa> 
baa)> 

raaffsr 
faaOor 
fiMkSor 
&a9ii) 



fftam 
kaasl 

IT 



flaask 



staalii) starling stser 



laart 



maast 

kaast 

gaastli 

blaart 

haasp 

maaj 

haaf 
laaf 
■taaf 

kaaf 

tjaaf 

aaftar 

raaftor 

laaftor 

Jaaft 

kraaft 

draafi 



haavist 
staav 
kaav 
kaav 



jaan 

baan 

aansor 



garlidc 
eharioek 
darling 
barley 

hearth 
lath 
path 
bath 

rather 
father 
farther 
farthing 

ass 

arse 

fasten 

caMle 

grass 

glass 

ask 

fiask 

baA 

last 



fast 

mast 

east 

ghastly 

hUut 

hasp 

marsh 

half 

laugh 

staff 

calf 

chaff 

after 

rcfter 

laughter 

shcft 

erafl 

! draught \ 
drq/« \ 

harvest 
starve 
oaXve 
carve 



yam 
bam 
answer 



Ofrlic 

dSorlin^ 

beerlic 

h6or]» 

Lett 
p»)» 

hrajKir 
fieder 
farpoT 
f^r]>lui^ 



fiBsUmian 
cactel 



glss 
Sscian 
flasoe 
ba)Mk 

/hlsest 

Jlatost 

^last 

(lAstaa 

(fiest 



kasta 
gsestlio 
blAst (6) 



m^rse 

half 

hl^hhaa 

BtBef 
(calf 
{kalfi 

cef 

aefter 

rsefter 

hlahtor 



crseft 
diaht 

hserfest 
steorfitt 

ceorfan 

geani 
bnm 
andswaru 



SECOND W0&D-LI8T. 



383 



aant 
Jaant 



haam 
Immuh 

Mk 

hAAk 

hftftkn 

laak 

SpAAk 

nuMk 

dMk 

paak 

Sftt 

haat 
bMt 
muMi 
kMt 

tMt 

MtJ- 

hMd 

hftinln 

jaad 

hftftp 
Jaap 



hoor 

•tMr 

BpMr 



Ml 

whaol 
fiMlon 

tfMl 
WMld 

bM> 



ant 
thaUnol 

arm 
aima 
harm 
harm 

ark 

hark ) 

heathen \ 

lark 

etark 

epark 

mark 

dark 

park 

hark 

art 

hart 

heart 

emart 

cart 

tart 

arch- 

hard 
harden 

Sford 

harp 
ikarp 



her 
$tir 
•pur 
were 

earl 
whirl 
furlong 
ekuH 
world 

earth 

worth 

mirth 

girth 

hirlh 



Ainott6 
soasl nftwihi 



elmeme 

liMurm 

beormA 

earo 

hercniaa 

Uweroe 

■tearo 

fpearoa 

mearo 

deoro 

peanoo 

(beoioan 

eari 

heorot 

heorto 

smeortan 

onBt 

teart 



feest 
daeat 

bOMt 



ibas 
>oasdi 



worst 

first 

duret 

hurst 



^irte 
thureday 



wyrreat 

i^n^sta 

donte 

Sberstan 
bonten 

fyn 
^Hres-dsg 



waejip worship weor^aoipa 



skeaf 
toaf 



haard 
har^na 

(geaid 

<g«rd 

hearpe 
soearp 



hire 
Btyrian 
•pura 
wAron (6) 

•oil 

hvirfla 

furlang 

oeorl 

womld 

eor)»e 

weor^ 

myig> 

SSL 
geoyrtt 



ean 

oonirt 

jaen 

10911 

■iaan 

Bpaan 

fban 

kaanl 

teon 

boan 



scurf 
turf 

swerve 

earn 
earnest 
yearn 
learn 

stem 

spurn 

jem 

kernel 

trnm 

hwm 



ionrf 
tarf 

BweoT&n 



geeamian 
eomeBt 
gaom 
leomiaii 

iitfrna 
Btj5ni 
gpuman 
feam 
oymel 
tomian 
beoman 



worm wymi 

waamwnd wormwood wermod 



^fitrther toay/t 



kaae 

lioatt 



worte 
eurse 
thirst 



eun 
Ionian 



oak 
imoak 

woak 

maaki 

]«ati 

)iao(ijn 

Jaat 

waai 

kootl 

toatl 

daat 

^/ 

haad 

haad 

haadl 

tJaaTU 

]iaad 

woad 



irk 
smirk 

work 

mirkg 

thirty 

thirteen 

shiH 

wort 

kirUe 

twite 

dirt 

ehur^ 

hireh 

herd 

heard 

hurdle 

chervil 

third 

word 



maadar murder 



baadn 



girdle 

bird 

hurden 



■me(a)roiaii 

Iweoro 
wyroan 
myroe 

Jiritig 

JvSoiSne 

■kyrta 

wyrt 

oyrtel 

tortla 

drit 

cirioa 

biroe 

heord 

ffoherde 

nyrdel 

oerfilla 

^dda 

word 

imyr^ran 
mor^r 
gyrdel 
bridd 
byr^en 



384 



HISTORY OF ENGLISH SOUNDS. 



Wj 

PJ 

Hj 
J>rij 

eij 

«ii 

81J 

• • 

"J 

n. 

WIJ 

fij 

frij 

flij 

nij 
mi] 

kij 

glij 
trij 

pij(kok) 

bij 

bij 

ijl 
hijl 

biji 

rijl 

sijl 

Btijl 

Btijl 

wMjl 

fiji 

mijl 

kijl 
dijl 
iijld 

m 

wijld 
fijld 

hij> 
rij> 

Dinij)? 

bikwij]) 

tij]> 

fliJB 

ijst 
iJBtar 

jiJBt 

lijst 
priJBt 



he 

y« 

lee 
lea 
three 
thee 
the 
$ee 
sea 
the 
we 
fee 
free 
flee 
fiea 
knee 
tne 
key 
gUe 
tree 

pea(eock) 
bee 
he 

eel 

heel 

heal 

reel 

seal 

ettel 

ileal 

wheel 

feel 

meal 

keel 

deal 

yield 

tihield 

ioield 

field 

heath 

wreath 

sheath 

beneath 

bequeath 

teeth 

seethe 

fleece 

geese 

east 

easier 

yeast 

least 

priest 



g« 
hlSo 

l6ah 

JnrSo 

\a 

Be 

B^(n) 

§& 

bSo 

w5 

feoh 

fleo(n) 
fl6ah 
cnSo 
mS 



glBo 

treo 

pSa 

bSo 

bSo(n) 

ffl (6) 

hSU 

hffilAn 

hrSol 

Beolh 

BtSle 

stelan 

hwSol 

foelan 

melu 

kJ9l 

d£l 

geldan 

Bceld 

gew^ldan 

feld 

BC£)» 

beneo]>an 
becwe])an 

t(£)> 

sSo)«li 

fiSOB 

Sast 
SaBtron 

gOBt 
ls&8t 

prSost 



wijzl 


weazel 


weBole 1 


whijz 


wheeze 


hwsBan (S) 1 


Bnijz 


sneeze 


fneoean 


fnjz 


freeze 


fr€o0an 


tijz 


tease 


tSsan 


tfijz 

% 


cheese 


cfiBe (S) 


pijz 


pease 


p€oBan 


bijzam 


besom 


besma 


^K 


lief 


ISof 


^j/. 


leaf 


ISaf 


biijf 


belief 


geleafa 


m 


thief 


|>9of 


fijf 


sheaf 


Bc6af 


ijvl 


evil 


yfei i 

efen 


ijyn 


even 


ijvniij 


evening 


&fen 


hijv 


heave 


hgbban 


birijv 


bereave 


berSafian 


lijv 


leave 


leaf 

laefan 


bilijv 


believe 


gele£ui 


Blijv 


sleeve 


BlSf 


wijV 


weave 


wefan 


wijvl 


weevil 


wifel 


fijver 


fever 


fefer \ 


klijv 


cleave 


K clSofan 
( difian 


ijvz 


eaves 


efeB 


lijn 


lean 


( hleonian 
jhlegne 


Bijn 


seen 


geaSne 


fijn 


sheen 


BoSne 


wijn 


ween 


woenan 


wijn 


wean 


w^nian 


• « 




( maenan 
( gemsene 


mijn 


mean 


kijn 


keen 


cdene 


klijn 


dean 


ol£ne 


kwijn 


queen 


CWOPA 


kwijn 


quean 


cwene 


grijn 


yreen 


groene 


-tijn 


'teen 


-tene 


bitwijn 


between 


betwSonan 


bijn 


been 


*gebeon 


bijn 


bean 


bean 


fijnd 


fiend 


{^nd 


sijm 


seem 


BC man 


sijm 


seam 


B6am 


Btijm 


steam 


Bt6am 


Btrijm 


stream 


stream 


glijm 


gleam 


gl^em 


tijm 


teem. 


teman 


tijm 


team 


team 


drijm 


dream 


dream 


bijm 


beam 


beam 



SECOND WORD-LIST. 



385 



rijk 
lijk 
lijk 

«ik 

■nijk 

spijk 

wijk 

wijk 

mijk 

tfjk 
bijkar 

bijkan 

ijt 

hijt 

sijt 

•wijt 

itrijt 

fijt 

whijt 

fijt 

flijt 

mijt 

mijt 
miJt 
grijt 

bijt 

bijtl 

bliji 

ijt/ 

rijtf 

rijtf 

lijtj 

bisijtf 

BpiitJ 

tijtf 

bijif 

brijtf 

brijtj 

bUjiJ 

hijd 

rijd 

rijd 

rijd 

lijd 

sijd * 

stijd 

Bpijd 

wijd 

fijd 

nijd 

nijd 

nijdl 



redt 

leek 

Uak 

teeh 

$neak 

8peak 

wed: 

weak 

meek 

ckeek 

heaker 

beacon 

eat 
heat 
seat 
eweet 
itreet 
theft 
wheat 
feH 
fleet 

meet 

meat 
mete 
greet 

heat 

beetle 

bleat 

each 

reach 

retch 

leech 

beseech 

epeech 

ieach 

beech 

breech 

breach 

bleach 

heed 

reed 

rede 

read 

lead 

feed 

steed 

ppeed 

weed 

feed 

need 

knead 

needle 



SSoan 
Sm 
rSo 
l6ao 
leka 
B<Bcaii 
mioan 
Bpreoan 
wica 
▼eik 
mjak 
cSaoe 
bikar 
bSacn 

eton 

hftto 

•Sti 

iwdete 

strftt (S) 

■cQte 

hwffte 

fOBt 

fleote 

igenuBtan 
iii«te(S) 
mfte 
meUn 
ffroetan 
I Matan 
b6ot 
bitol 
betel 

blertan (9) 
fflo 



hrscan 

Ivoe (B) 

bescecan 

ipneo (6) 

tffcan 

bdrce 

broMS 

bryoe 

bls€an 

hoedan 

hrSod 

rftd(6) 

rAdan (6) 

lAdan 

8fid(S) 

ttoeda 

Bp&dan 

w6od 

fc dan 

ned 

cnedan 

nftdl (S) 



mijd 

mijd 

krijd 

grijdi 

dijd 

bijd 

brijd 

blijd 

hijp 

rijp 

lijp 

Blijp 

iwijp 

Btijp 

rtijpl 

wijp 

kijp 

krijp 

tfijp 

djjp 



ior 

hiar 

hiar 

jiar 

riar 

liar 

siar 

imiar 

stiar 

spiar 

jiar 

jiar 

wiap 

wiari 

fiar 

niar 

miar 

giar 

tiar 

diar 

diar 

driari 

biar 

biar 

Uiar 

wiad 

biad 



hei 

• • 
joi 



meed 

mead 

creed 

greedy 

deed 

bead 

breed 

bleed 

heap 

reap 

leap 

sleep 

sweep 

eteap 

eteeple 

sheep 

weep 

keep 

creep 

cheap 

deep 



ie 



ear 

here 

hear 

year 

rear 

leer 

sear 

smear 

steer 

spear 

shear 

sheer 

wier 

weary 

fear 

near 

mere 

gear 

tear 

deer 

dear 

dreary 

beer 

bier 

blear 

weird 

beard 



iii<6d 
medn 
mAd(6) 
orSda 
griWjg :p) 
dSd (S; 
gebed 
broedan 
bid* dan 

hfSap 

reopan 

hl6a|an 

slApan (9) 

sw&pan 

BtSap 

itQpal 

scl&p (6) 

WGBpan 

ocepan 

orSopan 

o6ap 

d6op 



9are 
hSr 

gehSran 
gSar (gfr) 



ei 



hay. 
yea 



hl9or 

sSarian 

Nnero 

atSoran 

■pere 

iceran 

wkSr 

war 

wdarig 

n6ar 

na^ra 

gearwe 

tear 

dfior 

dfiore 

drSorig 

beor 

b*r(6) 

blere 

wyrd 

beard 



hfig 



C C 



386 



HISTOET OF ENGLISH SOTINDS. 



(b])rei 


bewray 


wrAgan 


neiv 


nave 


nafa 


lei 


lay 


{hdg 


neiv 
neivl 


knave 
navel 


cnafa 
nafola 


'Sei 


they 


>eir 


kreiv 


crave 


orafian 


eei 


my 


HCgMl 


geiv 


gave 


g«f 


slei 


slay 


dean 


greiv 


grave 


g^f 


Bwei 


stoay 


sreigja 


BteiYZ 


staves 


Bta&s 


wei 


teay 


weg 








wei 


vey 


w«ge(5) 


rein 


rain 


regn 


whei 


whey 


bwag 


lein 


lane 


lane 


nei 


nay 


nei 


lein 


lain 


gelegen 


nei 


neigh 


hn^tgan 


)>ein 


thane 


Jwgen 


mei 


may 


nuBg 


Blein 


slain 


BliBgen 


klei 


clay 


dfig 


Bwein 


sw€Un 


Bvein 


grei 


gray, grey 


dKg 


wein 


wane 


wanian 


dei 


day 


wein 


wain 


wsBgn 


plei 


play 


plegpan 


vein 

• 


vane 


&na 








mein 


mane 


mana 


eU 


ale 


ala 


mein 


main 


msegen 


eil 


ail 


eglan 
hftl 


krein 


crane 


cran 


beil 


hale 


twein 


twain 


twragen 


beU 


hail 


hmgl 
hei! 


drein 
bein 


drain 
bane 


drehnian 
bana 


sell 


mlt 


sal 


brein 


brain 


bnegen 


sell 


tail 


segl 


(tj]l)blein 


{chilPfdain 


blegen 


sneil 


anad 


snsegl 


eint 


am not 


eom nSwiht 


skeil 


secUe 


. Bcalu 








well 


wail 


weil&wei 


leim 


lame 


lama 


whdl 


whale 


hwsel 


Beim 


same 


sand 


neil 


nail 


n»gl 


Jeim 


shame 


Bcama 


naltingeil 


nightingale 


n^tegale 


neim 


name 


nama 


teil 


tale 


talu 


keim 


came 


QwOm 


teU 


iaU 


tsBgel 


geim 


game 


gamen 


deil 


dale 


diel 








peil 


paU 


p»gel 


eik 


ache 


acan 








eiker 


acre 


soer 


lei« 


lathe 


I9J, 


eikon 


acorn 


aeoem 


■weiS 


iwathe 


Bw^J^ian 


reik 


rake 


race 


bei« 


bathe 


ba^an 


leik 


lake 


laou 








Beik 


sake 


Bacu 


reiB 


race 


rftg 


Bleik 


slake 


slaoian 


weiBi 


waigt 


wsBstm 


sneik 


snake 


snaca 


weist 


waste 


woBstan 


steik 


stake 


staca 








Bteik 


steak 


Bteik 


beizl 


hazel 


hsBel 


Bpeik 


spake 


spneo 


reiz 


raiae 


reiia 


Jeik 


shake 


scaoan 


grei* 


grase 


ffrasian 
dsBgeseage 


weik 


wake 


(on)waoan 


deisi 


daisy 


we^kn 


waken 


(S)wsBcnian 


breizn 


brazen 


br»9en 


fleik 


flake 


flaki 


bleiz 


blaze 


blseee 


neikid 


naked 


naood 








meik 


make 


madan 


tjeifar 


chafer 


c»fer 


keik 


cake 


kaka 








kweik 


quake 


cwacian 


biheiv 


behave 


bebabban 


teik 


take 


tacan 


beivn 


haven 


h^fii 


beik 


bake 


baoan 


reiyn 


raven 


hrafn 


breik 


brake 


brtBc 


Jeiv 


shave 


flcafan 


breik 


break 


brecan 


• 




) wafian 

1 W«g (6) 








weiv 


ware 


eit 


ate 


fit 


weivor 


waver 


vafra 


eit 


eight 


sehto 



SECOND WOBD-LIST. 



387 



eii 


mfot 


«g»J» 


meer 


mare 


mfre 


belt 


hate 


bfttian 


keer 


eare 


earn 


leit 


laid 


Ute 


teer 


tear 


teraa 


skeit 


skats 


skata 


tjeeri 
deer 


eharg 


eeorig 


weit 


feeight 


gewibte 


dare 


dearr 


geit 

greit 

belt 


great 


geet 
great • 


peer 
beer 


pear 
bare 


pera 
tmradj. 


baU 


beitasft. 


beer 


bare 


hsaeprt. 


belt 


baU 


beita«6. 


i_ _ 


bear 


(bera 
( beran 


eit> 


eighth 


tehtojMk 


beer 


reid 


raid 


rSd 








leid 


lade 


bladan 




uw 




leid 

leidi 

leidl 

Rpeid 

Jeid 

weid 


laid 

lady 

ladU 

epade 

Made 

wade 


Ifgde 

blAfdige 

hlsBdel 

spade 

sceada 

wadan 


bnw 

mw 

Jnrnw 

Jnruw 

slnw 


who 

me 

threw 

through 

slew 


bwft 

brOowan 

)nr6ow 

^urb 

laog 


meid 

meid(n) 

kreidl 


made 

maid(en) 

cradle 


maoode 

XDflBffden 

eradol 


straw 

raw 

Jruw 


strew 

shoe 

shrew 


strQwian 

BOO 

Ber6awa 


breid 
bleid 


braid 
blade 


bregdan 
blied 


wow 
flnw 
kruw 


woo 
Jlew 
crew 


wOgian 

fl9ab 

crSow 


9 






kluw 


clew 


clfiowe 


eip 
beipni 


ape 
halfpenny 


MM 

bealf ^ning 


graw 
tnw 


grew 
too 


gr6ow 
t5 


skreip 
Bteipl 


sorape 


serapian 


taw 


two 


twft 


staple 


stapol 


trow 


true 


tr6owe 


Jeip 


thape 


skapa 


tfaw 
duw 


chew 


oeowan 


meipl 


maple 


mapnlder 


do 


d0(n) 


goip 


gape 


g»P* 


druw 


drew 


drOg 


teip 


tape 


teeppe 


bruw 


brew 


brtowan 


teiper 


taper 


taper 


Uaw 


bUw 


blSow 


neiber 


neighbaitr 


nSabgebOr 


stnwl 
sknwl • 


stool 
school 


stol 

BOOl 








tuwl 


tool 


tol 




•• 




puwl 


pool 


p5l 


eer 


^9^ 


Or 


niw]> 


ruth 


•br5ow> 


beer 


hare 


bara 


8nw)> 


sooth 


B0> 


beer 


hair 


bKr(6) 


ankaw]> 


uncouth 


uneQ)> 


0eer 
0eer 
■weor 


there 
their 
swear 


JriftrCS) 
Bw^an 


taw)> 

traw)> 

buwj> 


tooth 
truth 
booth 


tQ> 

tr9ow)> 

h6y 


sneer 


snare 


sneare 


■maw9 


smoothe 


smCjw 


steer 


stare 


starian 








steer 


stair 


stftger 


lawB(n) 


loos€{n) 


l^nsCna) 


Jeer 


spare 
share 


sparian 
scser 


guws 
rnwst 


goose 
roost 


brOst 


jeer 


share 


scam 








weer 


ware 


waer 


uws 


ooze 


wOb 


weer 


wear 


w^an 


uwd 


onset 


Osle 


wbeer 


where 


hw«r (e) 


Inwx 


lose 


(for)l6oBan 


feer 


fare 




tjaws 


choose 


o6osan 


feer 


fair 


ftsger 


bruwx 


bruise 


br6Ban 


(fijld)feer 


Jieldfare 


felofor 








(nAit)meer 


nightmare 


mare 

C 


'nwf 

c a 


roof 


brOf 



388 



HISTORY 01' ENGLISH SOUNDS. 



wuwf 


woof 


Owef 


tjuwzdi 


tueaday 


tiwes-dieg 


pruwv 


prove 


pr5fian 


njuwt 


newt 


^te 


Bnwn 


toon 


aSna 








Bwuwn 


twoon 


(ge)8w5gen 




Jao 




nuwn 


noon 


nCn 


judr 


your 


Sower 


muwn 


moon 


mSna 








buwn 
wuwnd 


boon 
wound 


b5n 
wund 


stja»d 


steward 


stigweard 


llUWID 


whom 


hwSm 




ou 




luwm 


loom 


ISma 






^ __ 


wuwm 


womb 


wamb 


ou 


0100 


figan 


fflawm 
dawm 


gloom 


glQm 


rou 


roe 


rft 
hrogn 


doom 


dom 






blawm 


Uoom 


blQma 


rou 


row 


rSw 
rOwan 


luwk(wgm) lukewarm 


wlacu 


lou 


lo 


lOea 
] hlOwan 






r6t 

wrOtan 

BcSotan 

gemSt 

bOt 


lou 


low 


ruwt 


root 


]7rou 


throe 


>ra 


Jnwt 

muwt 

bawt 


sKoot 
moot 
hoot 


yrou 

0OU 

sou 


throw 
though 

90 


jnrSwan 
Ji5h 

BWS 








sou 


tow 


sftwan 


rawd 

lawd 

fdwd 

mawd 

brnwd 


root 

lewd 

food 

mood 

brood 


rod 

l£wed 

fbda 

mod 

brod 


sou 

bIou 

bIou 

snou 

stou 

Jou 


sew 

doe 

slow 

snow 

stow 

show 


sSowan 

ds 

slSw 

snSw 

BtOwian 

BoSawian 


stawp 


etoop 


stQpian 


won 
fou 


woe 
foe 

flow 


wS 
ft 


whnwp 


hoop 


hwCpan 


frou 


fra 


dmwp 


droop 


drQpa 


flou 


fiSwan 








nou 


no 


n& 






• 


nou 


know 


cnRwan 




no 




mou 


mow 


mftwan 


mU9T 


moor 


m5r 


krou 


crow 


crftwe 
crawan 


















gou 


^0 


g«(n) 




jaw 




grou 


grow 


grOwan 


9 






glou 


glow 


gl5wan 


JUW 


y(m 


Sowj[»m. 


tou 


toe 


tft 


uw 

4 


yew 


Sow sb. 


ton 


tow 


tow 


ijnw 


ewe 


eowe 


antouod 


unioward 


*uiitOweard 


hue 


hSow 


dou 


doe 


dfi 


hjaw 


hew 


hSawan 


dou 


dough 


diff 


J'juwz 


thewe 


])SawaB 


bou 


bow 


boga 


spjuw 
fjuw 


fpew 
few 


spiwan 
fSawe 


blou 


blow 


blSwan 
bl5wan 


njuw 


new 


nSowe 








njuw 


knew 


cnSow prt. 


houl 


hole 


bol 


mjuw 


mew 


mffw 


houl 


whole 


hSl 


djuw 


dew 


dSaw 


houli 


holy 


hiaig 








)»oul 


thole 


>ol 


juwl 


ytde 


geoU 


soul 


soul 


sftwol 








Bwouln 


swollen 


swollen 


juwj> 


youth 


««>g«> 


Btouln 


stolen 


stolen 



SECOND WORD-LIST. 



389 



foal 


shoal 


flooltt 


kloav 


doce 


dofe 


foul 


foal 


foU 


kloavor 


clover 


dftfre 


noQl 


knoU 


enoU 


klonvn 


cloven 


clofen 


moul 


mole 


mSl 


grouv 


grove 


grif 


^ koul 


coal 


ool 


drouv 


drove 


drSf 


tool 


toU 


toll 








donl 


doU 


gedsl 


oon 


otm 


Sgen 


pool 
bool 


pole 


ES 


houn 


hone 


hftn 


boU 


onnli 


only 


snlio 


bonl 


hovl 


bolla 


loan 


loan 


Iftn 


bonlster 


hoUter 


bolster 


oloun 


alone 


allftna 


monltn 


molten 


molten 


stoan 


stone 


stftn 


konlt 


coU 


colt 


flonn 


Jloicn 


flogen 


konltar 


coulter 


cnlter 


moon 


moan 


mJensn 


bonlt 
onld 


hoU 
old 


bolt 
ftld 


groan 
droan 


groan 
drone 


crSnian 
drftn 


honld 


hold 


luddan 


boan 


bone 


bftn 


■onld 


Bold 


■aide 


wonnt 


wont 


gewanod 


Jocddw 


ehotilder 


Bcnldor 


wonnt 


ufUlnot 


wile nftwiht 


isr' 


l/M 


fftlod 








fold 


faldftn 


honm 


home 


bSm 


moald 


mould 


molde 


loom 


loam 


Um 


koald 


cold 


odd 


foam 


foam 


fSm 


goold 


gold 


gold 


koom 


comb 


csmb 


toald 


told 


tidde 








bonld 


hold 


bold 


oak 


oak 


fie 






• 


oakom 


oakum 


fioambft 


on> 


oath 


ft> 


Joak 


yohe 


geoo 


loa]> 


loath 


lftj» 


joak 


yolk 


geolca 


■lon)> 


Bloih 


idSw)> 


soak 


$oak 


Booian 


kwou^ 


quoth 
both 


ewwp 


smoak 


emoke 


smocian 


bon^ 


bftjnr 


stroak 


ttroke 


strioian 


tnmjy 


troth 


lrtow> 


spoak 


epoke 


(spsea 
\ Bproo 


lontS 
klou9s 


loathe 
elothee 


Ujrian 


spookn 
woak 


spoken 
woke 


gesprecen 
(on;wQo 








foak 


folk 


folo 








kzookr 


croak 


orioettaa 


mouBt 


mofi 


mnit 


toakn 


token 


tSoen 


gonst 


ghoei 


gist 


tjoak 
poak 
broak 


choke 


oeooiaa 


pouBt 


poti 


post 


poke 
broke 


poki 
brtBo 


hou 


kote 


hose 
(rose 
trts 


broakn 


broken 


gebrooen 


rou 


rote 


oat(B) 


oat 


-ftte 


Sons 


thoae 


^ 


roat 


wrote 


writ 


Q^fUQ 


frozen 


froren 


Jvoat 


throat 


Jnote 


noiu 


noee 


nosa 


smoat 


smote 




tfoM 


ehoee 


cess 


(o)float 


(a^eooe 


flot 


tjonzn 


ehoeen 


coren 


float 


float 


floUan 








moat 


mote 


mot 


ouf 


ocf 


nlf 


goat 


goat 


«*t 


loaf 


loqf 


hlftf 


groat 
boat 


groat 
boat 


^ 


onTor 


over 


ofer 








1 honv 


hove 


h5f 


road 


rode 


Hdprt, 


' bihouT 


behove 


behsfian 


road 


road 


rid«6. 


•tour 


etove 


stofe 


load(8toan) load{Mtone) 


Ud 


wouvn 


woven 


wefen 


stroad 


strode 


strid 


kony 


cove 


cofa 


woad 


woad 


wid 



390 



HISTOBT OF ENGLISH SOUNDS. 



gond 


goad 


g«d 


ho0 


hoaree 


hifl 


tood 


load 


tfidige 


kro8 


cross 


kroM 


bond 


bode 


bodian 


g«* 


gorse 


gor^t 


oboud 


abode 


Sbftd 


lost 


lost 


ffeloeod 
nroet 








frost 


frost 


oupn 
hoap 


open 
hope 


open 
hopa 


of 


off 


of 


roup 


rope 


rftp 


ofo 


often 


oft 


■oup 


soap 


sflpe 


kof 


cough 


oohhettan 


group 


grope 


grilpian 


trof 


trough 


trog 


ponp 


pope 


p^» 


dwof 


dwarf 


dweoig 








loft 


Iqft 


loft 








■oft 


soft 


■Ofte 









kroft 


croft 


croft 





awe 


•gi 








ho 


haw 


haga 


on 


awn 


9gn 
horn 


ho^n 


hawlkorn 


hagu)x)m 


hon 


hom 


ro 


raw 


hrSaw 


honit 


hornet 


hymetu 


lo 
auUo 


law 
outlaw 


laffu 
aUaga 


jon 
felon 


yawn 
forlorn 


geonian 
forloren 


J» 


thaw 


^wan 


^n 


thorn 


^om 






(gessBh 


■won 


sworn 


Bworen 


89 


saw 


iw«« 


Jon 


sliom 


■ooren 






I "t^e^ 


won 


warn 




tiro 


etraw 


■trSaw 


fon 


fawn 


fiegnian 


J^ 


ahaw 


Boaga 


mon 


mourn 


muman 


no 


gnaw 


gnagan 


kon 


oom 


oom 


mo 


maw 


maga 


gon 


gone 


g«g*a 


klo 


daw 


dawu 


ton 


torn 


toren 


dro 


draw 


dragan 


don 


dawn 


dagian 








bon 


bom{e) 


geboren 


el 


awl 


awel 








ol 


all 


all 


hom 


haulm 


halm 


hoi 


hall 


hall 


Bwom 


swarm 


Bwearm 


>iol 


thraU 


yrsa 


■torn 


storm 


■toim 


smol 


email 


■msl 


worn 


warm 


wearm 


stol 


Ma 


■tall 


femer 


former 


forma 


BtolwOt 


ftdlwart 


ft8Blwir>e 


kwom 


qualm 


owalm 


wol 


waU 


wall 




X 




wolnat 


walnut 


yalhnot 


hok 


hawk 


hafoc 


fbl 


fall 


fallan 


Btok 


stalk 


■talnian 


kol 


eaU 


oeallian 


■tok 


stork 


Btoic 


krol 


crawl 


krafla 


wok 


walk 


waldan 


«1. 


gall 

44 ^ 


gallA 


fek 


fork 
mawkish 
chalk 
baulk 


forca 


g»ld 

older 

joldemen 


0dled 

alder 

alderman 


gallede 

aler 

aldennann 


mokif 

tJok 

bok 


ma)>k 
cealc 
bale 


ro]> 


wrath 


wrsbbo 








ro> 


wroth 


wraj> 


oger 


auger 


nafogSr 


■wo)>i 


awarthy 


Bweart 








fb> 


forth 


for> 


ot 


aught 


ftwiht 


fe> 


fourth 


f^r^ 


ot 


ought 


fthte 


firo> 


froth 


fro^a 


rot 


wrought 


worhte 


no> 


north 


nor|> 


>ot 


thought 


))5hte 


mo)> 


moth 


mo^^ 


]ywot 


thwart 


J>Tert 


klo> 


doth 


oU> 


■ot 


sought 


B^hte 


bro)> 


broth 


brob 


■loter 


slaughter 


■lahtr 






^ 


Jot 


short 


Boort 


hofi 


horee 


hors 


wot 


wart 


wearte 



SECOND WOED-LIST. 



39 » 



fdt 
foil 
fbtijn 
fotnait 

not 

tot 

dotor 

pot 

bot 

brot 

otjdd 

•kotj 

odn'l 

hod 

lod 

■od 

■wod 

wod 

fod 

afod 

bod 

brod 

todz 

wop 



or 

or 

or 

box 

hor 



water 
fought 

forty 

ftmrteen 

forinighl 

naught 

taught 

daughter 

port 

bought 

hrought 

orchard 

•eorch 

ordeal 

h)ard 

lord 

9Word 

eward 

ward 

ford 

afford 

hoard 

broad 

towards 

warp 



or 

oar 

ore 

hoar 

ukore 



bor(b)»imd horthovmd 



jor 
ror 
lor 
■or 



skor 

for 

bifor 

for 

flor 

nor 

ntor 

gor 

dor 

bor 

bor 



ai 



yore 

roar 

lore 

tore 

twore 

teore 

for 

btfore 

four 

^oor 

nor 

more 

gore 

door 

bore 

hoar 



t 

eye 



weter 

fiaeht 

ftowertig 

f^towertCne 

ftowertfine 

n^ht 
nAwiht 
t»hie 
dohtor 
port 
bohte 
brObto 
ortg«Md 
skorpn* 

ordsl 

herd 

Ufiford 

Bweord 

Bweard 

weard 

ford 

g6for)nAii 

bord 

brftd 

tSweardM 

twearp 
weoipAii 



Sr 

Ora 
hSr 
h0re 

hArehQne 
gSam 
rsrian 
lir 
■fa- 
twOr 
■cora 
for 

beforan 
f^wer 
fl<Sr 

nShweJper 
mfae 
Jgor 

dura 
iber 
i borian 

bir 



10 

«age 



ai 
hai 
bai 
rai 

lai 

Ui 
W 



dai 
■kai 
Btai 
wbai 

flat 

nai 

tai 

dai 

dai 

drai 

bai 

idlond 

BtaU 

whail 

wluul(st) 

fail 

(di^faU 

mail 

pail 

waild 
maild 
t/aild 

aiVar 

raiS 

laifS 

■aiO 

nai'Sar 

taiff 

blaiS 



aye 
high 
hie 
rye 

lie 

Ige 

thigh 
tigh 

dy 
eky 
etye 
fthy 

Af 

nigh 

tie 

dye 

die 

dry 

buy 

itland 

ttile 

while 

whiltt 

JUe 

{de)file 

tiU 

pHe 

wild 
mild 
ehiid 

eiiher 

wriths 

lithe 

teythe 

neither 

tithe 

bUthe 



au 


tee 


lU8 


lice 


)irais 


thrice 


mais 


mice 


twaia 


twice 


knist 


chritt 


raia 


rite 


waiz 


wite 



Uuf 

waif 

naif 



Ufe 

wife 

kni/e 



9H 

heah 
higian 

Uiogan 

) l6ogan 

ISaff 

|«oh 

■loan 

■loBg 

Bkf 

Btiga 

bwy 

\ flSogaa 
nSah 

Mgan 

dSgaa 

djyja 

dr/ga 

lA 

bjogan 

flgland 
■tigal 
wbU 
^-whlle-^a 

fa 

fflan 

tigola 

mil 

ipa 

jpil 

wUda 
milda 
Gild 

fig]>ar 

wrijian 

n)>e 

Bijw 

*niigj>er 

t8ogo]naii 

bU>a 

is 

lya 

)xriwa 

myi 

twiwa 

orut 

aiiaan 

iwis 
wiaa 

llf 

wif 

enif 



392 



HISTOEY OP ENGLISH SOUNDS. 



aiYi 
liaiv 

)iraiy 

faiv 

Jraiv 

dfliv 

draiv 

lain 

«ai(n) 

Bwain 

Jain 

jrain 

wain 

whain 

nain 

mu(n) 

kain 

twain 

pain 

brain 

haind 

bihaind 
raind 
waind 
faind 

kaind 

sraind 

baind 

blaind 

raim 
raim 

laim 

klaim 
taim 

laik 

Btradk 

Jraik 

daik 

paik 

halt 

ndt 

ndt 

rait 

alait 

lait 

lait 

Bait 

Blait 

slait 

smait 

wait 



ivy 

hive 

alive 

thrive 

Jive 

$hrive 

dive 

drive 

line 

thine, thy 

Wfine 

ahine 

ehrine 

wine 

vhine 

nine 

mine, my 

hye 

twine 

pine 

hrine 

hind 

behind 
rind 
wind 
find 

hind 

grind 

bind 

blind 

rhyme 
rime 

hme 

climb 
time 

like 

ttrike 

shrike 

dyke 

fxke 

height 

right 

write 

Wright 

alight 

Ugh 

light 

lightning 

sight 

digkt 

deight 

smite 

wight 



Ifig 

hyf 

on life 

Jnifaak 

fife 

■ciifan 

dyfan 

dri&n 

line 

)an 

■win 

Bcinan 

Borin 

win 

hwinan 

nigon 

min 

oy 

twin 

ipinian 
pin(tr6ow) 

bryne 
ihind 
{ hiwa 

behindan 

rind 

windan 

findan 
Jgocynd 
(gepynde 

ffrindan 

bindan 

blind 

rim 

brim 
lljm 
(linden 

tima 

gelic 

Btrioan 

Bcric 

die 

pic 

reht 

wiitan 

wyrhta 

alihtan 

leoht 'levis' 

leoht 

ISgetu 

gesibp 

Bleht 

•lag)' 

Bmltan 
wiht 



whait 

fait 

frait 

flait 

flait 

nait 

nait 

nudt 

mait 

kait 

tait 

dait 

plait 

bait 

brait 

blait 

raitjds 

aidl 

haid 

raid 

Baid 

Blaid 

Btraid 

waid 

fraidi 

glaid 

taid 

taidii^z 

tjud 

praid 

baid 

braid 

braidl 

raip 

Bnaip 

waip 

graip 

paip 



haiar 

Bpaiar 

Jaiar 

waiar 

faier 

maiar 

taiar 

braiar 

ua-n 



white 

fig^ 

fright 

flight 

flight 
night 
knight 
mite 

might 

kite 

tight 

dight 

plight 

bite 

bright 

blight 

righteous 

idU 

hide 

ride 

side 

slide 

itride 

wide 

friday 

glide 

tide 

tidings 

chide 

pride 

bida 

bride 

bridle 

ripe 

rniipe 

wipe 

gripe 

pipe 



ai9 



hau 
Oau 



hire 

spire 

shire 

wire 

fire 

mire 

tire 

briar 

iron 



how 
ifwu 



hwit 

feohtan 

fjrrhto 

flyht 

flyht'flee'ing' 

n^ht 

cniht 

mite 

imfht 
msbte 
cyta 
|)eht 
dihtan 
pliht 
bitan 
beorht 
bl^)» 
rehtwis 

idel 

hid 

h/d 

hydan 

ridan 

side 

Blidan' 

stridan 

wid 

frlge-deeg 

glidan 

tid 

ti))endi 

<ddan 

pryte 

bidan 

bryd 

bridels 

ripe 

snite 

wipian 

giipan 

pipe 



hyran 

spir 

Bcir 

^nr 

fyr 

myr 

teonan 

brBre 

iren 



t 



au 



bo 



SECOND WOfiD-LlST. 



393 



■laa 
nau 

kan 

trau 

plan 

Daa 

ban 

bran 

aol 
faol 
faol 
kaol 

8an)> 
maa)» 

hauB 
lans 



eow 

dough 

now 

eow 

trow 

plough 

bow 

hough 

brow 

owl 
fowl 
fiml 
eowl 

south 
mouth 

house 
louKe 



mans mouse 

(tit)mai]B {tit)mou9e 



BUgU 

•log 
no 

Iktlga 
trCowian 
ploh 
bOgan 

bro 
Qle 

ongle 
0Q> 

has 

iQs 

mai 

mS86 



jMtnsond thousand ^Qiend 
drauzi drowsy drOsian 



taan town 

dann down 

braun brown 

graonal groundsel 

haimd hound 

boehaund horehowkd bire-htkne 



tan 

iofdane 
dOn'foathen' 
bran 

gnndeswilge 
hand 



■annd 

waond 

fannd 

grannd 

pannd 
bannd 

aut 

wi9ant 

lant 

kUnt 

danti 

drant. 

abaut 

laud 

fraud 

kraud 

klaud 

prand 



90und 

wound 
found 
ground 

pound 
bound 

out 

without 

lout 

dout 

doughty 

drought 

about 

loud 

shroud 

crowd 

doud 

proud 



isund 
geiund 
ffewunden 
fnnden 
grund 

Ipnnd 
punian 
ibunden 
bain 

at 

wi)Atan 

latan 

olat 

dyhtig 

draga)) 

ymbatan 

hlad 

icrad 

cradan 

dad 

prat 



au9 



anar 
■audr 
fauar 
iMbUOr 



our 
sour 
shower 
bower 



Ore 
sar 
scar 
bar 



oi 



boQ 



boa 



byle 



INDEX TO FIRST WORD-LIST. 



a artie. 1438 
Abbot 433 
abode 147 a 
about 1988 
above 997 
ache 373 
acknowledge 

1416 
aoom 375 
acre 374 
adder 1587 
addled 384 
»d«e 385 
after 180 
afford 1 191 
afloat 1330 
again 343 
ail 903 
ajar 670 
alder 64 
alderman no 
ale 6a 
alight 19 [I 
alive 1874 
all 76 
alms 96 
alone 1431 
also 84 
am 856 
amidst 638 
among(Bt) 310 
an 1438 
anchor 186 
ancle 187 
and 343 
anent 799 
anger, 199 
angle 300 
anon 1430 
answer 343 
ant 1534 
anthem 340 
anvil 339 
any 1513 
ape 40a 
apple 41a 
arch- 39 
are i 
ark 38 
arm 34 



arrow 33 
art 45 
as 84 
ash 138 
ashes 139 
ask 1398 
aspen 153 
ass 136 

»t 357 ^ 
ate 1570 
auger 171 
aught 1401 
awe 331 
awl 156 
awn 333 
axe 307 
axle 308 
aye 889 



back 300 
bad 1551 
bade 399 

bag 354 
bait 8b 918 

but vb 919 

bake 301 

ban 338 

bane 336 

bang 3X4 

bank 198 

bare adj 17 

hare prt 18 

bark sb 43 

bark vb 699 

barley 31 

barm 695 

bam 693 

(wheel)barrow 

38 

barrow 704 

bass 32 

bask 135 

bast 153 

bath 133 

bathe 134 

batten 376 

baulk loa 

be 1763 

beacon 1739 

bead 953 



beaker 585 
beam 1733 
bean 1717 
bear $h 665 
bear vb 666 
beard 57 
beat inf 1741 
hetktprt 1 831 
beck^on) 1645 
bed 957 
bedridden 636 
bee 1763 
beech 1644 
been 1815 
beer 1769 
beetle 630 
beetle 1665 
before 1183 

beg 954 
began 335 

begat 375 

begin 536 

begotten 939 

begun 1037 

behave 159 

behest 1493 

behind 543 

behove 3076 

belch 753 

belief 17 14 

believe 1634 

beU733 

bellow 759 

bellows 758 

beUy 758 

belt 764 

bench 816 

bend 854 

beneath 781 

bequeath 78a 

beineave 17x0 

bereft 1711 

berry 705 

beseech 1643 

besom 784 

best 946 

better 941 

between 181 4 

betwixt 594 

bewray 1655 

beyond 853 

bid 639 



bid(den) 635 
bide 1933 
bier 1560 
bill 438 
bill 443 
billow ma 
bin 538 
bind 555 
birch 430 
birth 1 104 
bishop 477 
bit 617 
bitch 593 
bite 1 93 1 
bitten 618 
bitter 619 
black 303 
bladder 1593 
blade 400 
blast 1568 
blase 135 
bleach 1530 
blear(eye(d) 

1610 
bleat 1 581 
bled 1678 
bleed 1675 
blend 855 
bless 1679 
blew 1806 
blight 877 
blind 556 
bliss 473 
blithe 1854 
blood 3143 
bloom 3097 
bloom 3098 
blossom 3065 
blot 1333 
blow 141 7 
blow 3073 
blush 1 131 
boar 1378 
board 1317 
boat 1463 
bode 1337 
body 1338 
boil aoi3 
bold 1x6 
bole 1337 
bolster 133a 
bolt 1339 



bond 354 
bone 1 441 
book 3108 
boon 3089 
boot 3X38 
booth 3054 
bore 18 
bom(e) 1 188 
borough 975 
borrow 1309 
bosom 3061 
both 1393 
bottom 1337 
bough 31 10 

bough 3X33 

bought X305 
bound 1046 
bound 1975 
bowsb 131 1 
bow vb 1986 
bower 1948 
bowl 1331 
box 1303 
braid 931 
brain 351 
brake 303 
bramble 1639 
brand 355 
brass 133 
braeen 134 
breach 1153 
bread 1746 
break 869 
breast 1785 
breath 1564 
breathe 1565 
breech 1646 
breeches 1647 
breed X674 
brethren 161 
brew 1805 
briar 1609 
bridal 63 
bride 3030 
bridge 1160 
bridle 1933 
bright 700 
brim 1150 
brimstone 

1 136 
brine 303a 
bring 533 



INDEX TO FIEST WOKD-LIST. 



brittle 1084 
broad 1473 
broke 30a 
broken 1286 
brood 2141 
brook vb 1981 
brook «6 2109 
broom 205^ 
broth 1245 
brother 2055 
brought 21 14 
brow 1943 
brown 1976 
bruise 1620 
book 1059 
build 1 1 15 
bulk 986 
bull 977 
bullock 98a 
bundle 1147 
burden 1078 
burial 1097 
bum 694 
bunt in/ 683 
burst j^^c 1196 
bury 1096 
busk 1942 
bustle 995 
busy 1 1 17 
but vb 1326 
but ej 1993 
butter 1007 
buttock 1069 
buy 2007 
by 1838 



cake 296 
<«lfo3. 
calf (of 1^)94 
call 51 
callow 90 
came 2091 
can v6 233 
can«6 234 
candle 252 
cannel (coal) 

cap 417 
can 15 
cart 372 
carve 686 
cast 150 
castle 151 
cat 380 
(oock)ohafer 

173 
chaff 172 

chaffer 1751 



chalk loi 
chapman 175a 
char 670 
charlock 673 
chary 16 
cheap 1750 
cheek 1728 
cheese 1^67 
chervil 687 
chest 791 
chew 1798 
chicken 578 
chide 1929 
child 452 
children 453 
chill 717 
chillblain 912 
chin 535 
choke 1282 
choose 1783 
chose 1697 
chosen Z185 
Christ 1861 
christen 1863 
Christendom 

1862 
Christmas 

1864 
churl 07a 
cinder 548 
dad 1492 
dap 418 
datter 373 
daw 157 
day 1536 
dean 1521 
deanse 1523 
deave 807 
deave 1812 
deft 1128 
dew 1800 
diff492 
dimb 566 
ding 522 
dip 645 
dip 1173 
dot 1329 
doth 1390 
dothee 1391 
doud 1999 
clout 1992 
dove 1267 
doyen 1268 
dover 1423 
duck 1297 
dung 1012 
duster 1124 
ooal 1225 
cobweb 1351 
cock 1294 



(oom)cockle 

"95 
cod 1340 
odd 114 
colt 1238 
oomb 272 
come 105 1 
comdy 1052 
cook 2105 
cool 2045 
copper 1343 
com 1 199 
cot 1324 
cough 1300 
could 1955 
coulter 985 
cove 1266 
cow sb 1941 
oow vb 1^84 
cowl 106^ 
oowslip 1 172 
crab 425 
crack 297 
cradle 395 
craft 183 
cram 267 
cramp 269 
crane 224 
crank 196 
crave 174 
crawl 1^5 
creed 1073 
creep 1836 
cress 786 
crow 1799 
cringe 521 
ori|^le 1 1 71 
crisp 488 
croak 1450 
crockery 1296 
croft 1273 
crook 2106 
crops6 135a 
crope6 1353 
cross 1254 
crow vb 1413 
crow «& 1414 
crowd 1998 
crumb 1053 
crumple 1055 
crutch 1 154 
cud 633 
cuff 998 
cunning 1026 
cup 1074 
oune g6$ 
outtle(fiBh) 

1072 



daisy 347 
dale 75 
daro 19 
dark ^8 
darling 1767 
daughter 1304 
dawn 348 
day 346 
dead 174R 
deaf 1716 
deal 1485 
tiear 1766 
death 1695 
deed 1591 
deem 1638 
deep 1837 
deer 1765 
deft 184 
delve 742 
den 842 
depth 1684 
devil 1813 
dew 1707 
did 1168 
die 898 
dight 597 

din 437 
dim 564 

din I 135 

dint 1 1 42 

dip 1174 

dirt 616 

dish 476 

distait485 

ditch 1910 

dive ao2i 

diny 1116 

do 2037 

dock 1298 

doe 1367 

dog 1313 

dole 1384 

done 2088 

•dom 2095 

doom 2094 

door 963 

dot 1331 

doih 2053 

dough 1456 

doughty 1060 

down s6 1971 

dowaadv 1972 

down ' feath- 

ers' 1973 

draft 320 

dr*g349 
drain 313 

drank 197 



395 

draught 320 
draw 349 
dread 1592 
droam 1722 
dreary 1768 
dregs 928 
drench 815 
drow 2120 
drift 502 
drink 511 
drip 2032 
drive 1884 
driven 494 
drone 1440 
droop 2003 
drop 1344 
dross 1253 
drought 1985 
drove |ir^ 1425 
drove «6 1426 
drown 1002 
drowsy 1962 
dry 2024 
drunk(en) 

lOOI 

dull 1226 
dumb 1057 
dun 1029 
dung 1014 
durst 966 
dust 1965 
dwarf 703 
dwdl 734 
dwindle 1896 
dye 1658 
dyke 1909 

e 

each 1487 
ear 1688, 9 
earl 671 
earn 31 
earnest 689 
earth 674 
earwig 606 
east 1698 
easter 1699 
eat 929 
eaves 797 
ebb 960 
edge 922 
ed 1 561 

e(K9a3 
eight 314 

dghth 315 

dther 1537 

eke vb 1640 

eke (2/ 1724 

dder 765 



396 



HISTOEI OF ENGLISH SOUNDS. 



elder (tree) 

720 
eldest 765 
eleven 847 
elf 737 
ell 744 
elm 745 
else 719 
embers 913 
emmet 1524 
empty 1525 
end 846 
England 818 
English 819 
enough 2 119 
ere 1480 
errand 1553 
even 798 
eTen(ing)i57o 
ever 1506 
every 1507 
evil 1 1 25 
ewe 795 
eye 1736 
eyot 1650 

f 

fain 336 

fair 335 
fall 80 
&II0W88 
false 84 
fSui 231 
far 668 
hre 13 
farther 964 

farthinff 1771 
fast 140, 7 
fasten 147 
fat 1544 
father 393 
fathom 120 
fawn 337 
fear 1558 
feather 779 
fed 1677 
fee 878 
feed 1670 
feel 1615 
feet 1662 
fell sb 728 
fell inf 729 
feXiprt 1776 
felloe 757 
fellow 16 1 6 
felly 754 
felt 761 
fen 838 
fern 32 



ferry 658 
fetch 874 
fetter 931 
fever 8oiS 
few 1706 
fickle 575 
fiddle 458 
field 769 
fieldfare 716 
fiend 1816 
fifth 1882 
fifty 1885 
fight 88^ 
file T84O 
(de)iile 2011 
fill 1 1 10 
film 444 
filth 2013 

fin 534 
finch 509 

find 553 

finger 520 

fire 2009 

first 1083 

fish 474 

fist 2018 

five i88z 

£1624 

flake 292 

flask 143 

flat 371 

flax 311 

flea 1686 

flee 1758 

fleece 1781 

fleet 1830 

flesh 1497 

flew 1738 

flicker 576 

flight 1 156 

flint 540 

flit 1 167 

flitch 589 

float 1 32 1 

flock Z293 

flood 2x36 

floor 2041 

flow 2070 

flown 1307 

flatter 1322 

fly v6 1824 

fly tb 1825 

foal 1224 

foam 1445 

fodder 2135 

foe 1361 

(Bheep)fold 72 

fold 113 

folk 1234 

follow 1236 



food 2134 
foot 2120 
for 1182 
ford 1 2 16 
forehead 1709 
foremost 1202 
fork 1204 
forlorn 11 76 
former 1 201 
forth 1 1 90 
fortnight 1795 
forty 1796 
foster 2063 
found 1042 
fonght 318 
foul 1 95 1 
four 1793 
fourteen 1794 
fourth 1770 
fowl 1062 
fox 1302 
foxglove 1269 
free 1757 
fret 370 
fret 932 
freeze 1780 
French 813 
fresh 681 
Friday 601 
friend 18x7 
fright 1092 
fro 1362 
frog 1312 
from 261 
fix«t 1255 
froth 1244 
froEen 1 184 
full 979 
fuller 980 
furlong 1949 
farrow 974 
farther 964 
furxe 108 1 



g 

gabble 426 
gadfly 401 
gall 82 
gaUed83 
gallow(8) 104 
game 263 
gander 253 
gang 212 
gannet 225 
gape 410 
garliok 1377, 

1727 
gate 374 
gather 396 



gave 177 
gear 27 
geese 16x9 
get 938 
ghastly 1503 
ghost 1400 
giddy 634 
gift 500 
gild 1114 
girdle 1103 
girth 678 
give 808 
glad 398 
glass 131 
glacier 132 
gleam 1526 
glee 1760 
glide 1930 
glisten 470 
glitter 615 
gloom 2093 
glove 2079 
glow 2072 
gnat 381 
gnaw 344 

go 1364 
goad 1470 
goat 1462 

god 1335 
gold 1243 
gone 1438 
good 2140 
goose 2059 
gore X186 
gore 1376 
gorse X195 
goshawk 2060 
gospel 1335 
gossip 648 

got 375 
grass 129 

grave 178 

gray 1575 
graze 130 
great 1740 
greedy 1590 
green 1633 
greet 1664 
grew 1802 

groy 1575 
grim 563 

grin 841 

grind 564 

grip 643 

gripe 1936 

grisly i860 

grist 1865 

gristle ^83 

grit 1x65 

groan 1439 



groat 1325 
grope 1477 
ground «6 1043 
ground ptc 

1044 
groundsel 447 
grove 1424 
grovel 1968 
grow 2071 
guest 793 
guUd 454 
guilt 1113 
gum 2092 
gust 994 
gat 106S 



had X85 

hail 324 

hail! 903 

hairi554 

hale 1379 

half 91 
hp.n 77 

hallow 1382 
ha(u)lm 97 
halt 106 
halter 95 
ham 264 
hammer 256 
hand 244 
handle 245 
hang aoi 
hank 188 
(mis)hap 414 
hard 49 
harden 50 
hare a 
hark 161 1 
harm*35 
haip 58 
harry 701 
hart 650 
harvest 29 
has 179 
hasp 154 
hat 377 
hate 358 
hath 179 
have 158 
haven 160 
hawk 161 
hawthorn 323 
hay 165a 
haxel 127 
he x6oo 
head 1708 
heal 1483 
health 1484 



INDEX TO FIRST WORD-LIST. 



hmp 1747 
heur 1600 
hewd i6ja 
hearken 161 1 
heart 706 
hearih 675 
heat 1539 
heath 1488 
heathen 1489 
heave 801 
heaven 803 
heavy 80a 
hedffe 890 
heed 1667 
heel 161 3 
heifer 1731 
height 1648 
held 1777 
hell 721 
helm 747 
helmet 746 
help 771 
hem 857 
hemlock 1 148 
hemp 315 
hen 835 
hence 8a8 
her 427 
-herd 431 
herd 709 
here 1605 
herring 3 
hew 1700 
hid 2031 
hide 1933 
hide 3028 
hide vh 2039 
hie 191 3 
high 1730 
hill 1 108 
hilt 448 
him 557 
hind 543 
hind 1867 
hinder 544 
hip 1 1 70 
hips 1834 
hire 30o8 
hi8 ^64 
hit 630 
hither 635 
hive 3030 
hoar 1369 
hoard 12 14 
hoarse 1393 
hold III 
hole 1 2 19 
holiday 1381 
hollow 1235 
holly 1 3 30 



(holly)hock 

1387 
holy 1380 
home 1443 
hone 1435 
honey 1015 
honeysnokle 

-hood 1464 
hood 2129 
hoof 3074 
hook ^099 
hop 1346 
hope 1343 
horehound 

1370, 1969 
horn 1 197 
hornet 1085 
hone 1 194 
hoee 1347 
hot 1458 
honnd 1034 
honse* 1958 
hooflewifei964 
hove 3075 
hovel 1 361 
how 1938 
hue 1789 
hoik 984 
hundred 1035 
hung 1637 
hanger 1003 
hnnt Z030 
hurdle iioi 
hoflband 1966 
hoflsy 1965 
hustingB 1963 



I 568 
ice 1855 
icicle 868 
idle 1933 

if 493 

511439 
imp 565 

inprp 534 

in adv 530 

inch 1 1 29 

inn 530 

irk 1088 

iron 1839 

18463 

island 165 1 

it 607 

itch 590 

ivy 1873 



k 

keel 718 
keen 1631 
keep 1683 
ken 840 
Kent 844 
kept 1685 
kernel 1086 
kettle 936 

W 1535 

kid 459 
kiU 730 

kiln 1 106 

kin I I 39 

kind 1 1 44 

kine 3006 

king 1 134 

kiss 1118 

kitchen 1153 

kite 3036 

kith 3014 

kirtle iioo 

knave 176 

knead 951 

knee 1759 

knell nil 

knew 1 801 

knife 1883 

knight 888 

knit 1 164 

knock 1383 

knoll 1239 

knot 1330 

know 1415 

knowledge 

1416 



lack 377 
ladder 1546 
lade 386 ' 
ladle 387 
lady 1511 
laid 930 
lain 909 
lair 900 
lake 278 
lamb 270 
lame 357 
lammas 1433 
land 346 
lane 316 
lank 190 
lap vb 404 

1*P4I5 
lappet 415 

lapwing 403 

lark 1403 



last < load' 141 
last 8pl 361 
last 1399 
last vh 1499 
latch 304 
late 359 
later 360 
latest 361 
lath 378 
lathe 118 
lather 1694 
latter 360 
laugh 879 
laughter 316 
law 326 

lay 891 
lea 1733 
lead vb 1545 
leads6 1743 
leaf 1713 
leak 863 
lean vb 839 
lean adj 151 5 
leap 1748 
learn 690 
least 1498 
leather 774 
leave vb 1508 
leave sb 1713 
led 1548 

1«« 1753 
leech 1 571 

leek 1736 
leer 1764 
leftfu{; iia6 
left ptc 1509 
leg 934 
lemman 181 o 
lend 1515 
length 830 
lent 1516 
Lent 817 
less 1496 
-less 1696 
lest 1494 
let ' hinder ' 

943 
let 1577 
let prt 1659 
lewd 1504 
lice 3015 
lick 584 
Ud638 
lie 605 
lie ' meniiri ' 

1833 
lief 1809 
life 1873 
lift II 27 



397 

light 1821 
light * levis * 

1833 
lightning 1653 
like 1904 
limb 559 
lime(tree) 547 
lime 1903 
limp 860 
linch(pin) 

1131 
linden 547 

line 1886 

linen 1887 

link 810 

linnet 535 

linseed 1899 

lip 644 

lisp 486 

list 1 133 

list<listen' 

1133 

listen 1 1 19 

listless 993 

little 3035 

lithe 1850 

live 647 

liver 489 

lo 3103 

loaf 1430 
loam 1443 
loan 1514 
loath 1387 
loathe 1388 
lobster 1347 
lock 1378 
lock (of hair) 

1389 
lode 1467 
loft 1373 
long 30 3 

look 3IOI 

loom 3090 
looBe(n) 3057 
lord 143 1 
lore 1373 
lose 1779 
lost 1349 
lot 1316 
loud 1996 
louse 1959 
lout 1991 
love 996 
low adj 1^55 
low vh 3008 
lukewarm 

391 
lung 1005 
lust 993 

ly« '737 



398 



HISTORY OF ENGLISH SOUNDS. 



m 

mad 1550 
madder 394 
made 295 
maid(en) 356 
main 341 
make 294 
mallow 89 
malt 108 
man 33a 
mane 22a 
many 223 
maple 409 
mar 669 
(night)mare 

mare 660 
mark 4 a 
marrow 44 
marsh 682 
mash 144 
mass 137 
roast 148 
mast 149 
match 876 
mattock 379 
maw 340 
mawkish 126 
may 339 
me 1603 
mead 950 
mead 1589 
meadow 1589 
meal 735 
meal 1563 
mean vb 1517 
meanflkf; 1520 
meant 15 19 
meat 934 
meed 1673 
meek 1819 
meet aflj 1 580 
meet vb 1663 
melt 763 
men 839 
mere 659 
merry 1094 
met 1666 
mete 935 
(Bea)mew 1 569 
mice 2015 
middle 637 
mid^ II S9 
midst 638 
might prt 319 
might sb 887 
milch 446 
mild 451 
mildew 434 



mile 1847 
milk 750 
mill 1 105 
mind 11 43 
mine 1894 
mingle 824 
minnow 1132 
minster 11 40 
mint 541 
mint 1133 
mire 2010 
mirky 1090 
mirth 1095 
miss 471 
mist 481 
mistletoe 48 a 
mite 1920 
mix 475 
moan 1518 
mole 1383 
molten 1237 
Monday 2087 
-monger 211 
monk 102 1 
month 2086 
mood 2137 
moon 2085 
moor 2042 
moot 2127 
more 1375 
morning 1208 
morrow 1207 
moss 1252 
most 1503 
mote 1323 
moth 1246 
mother 2139 
mould 1242 
mourn 971 
mouse 1 96 1 
month 1954 
mow 141 2 
much 577 
muck 1 151 
mugwort 1065 
mxadersb 1079 
murdert:i6ii93 
mussel 989 
must sb 993 
must rb 2064 
my 1894 

n 

nail 338 
naked 203 
name 262 
nap 416 
narrow 25 
naught 1 410 



nave 169 
nay 897 
navel 170 
near 1691 
neck 875 
need 167 1 
needle 1588 
neigh 1534 
neighbour 

»735 
neither 1538 

nest 790 

net 945 

nether 780 

nettle 933 

never 1510 

new 1797 

newt 800 

next 1649 

nib 962 

nigh 1734 

night 886 

nightingale 73 

nimble 561 

nine 602 

no 1363 

none 1429 

noon 2084 

nor 1419 

north 1 193 

nose 1250 

nostril 1251 

not 141 1 

nothing 1433 

now 1940 

numb 1050 

nun 102^ 

nut 1066 



o»^737 
oak 1446 

oakum 1447 

oar 1368 

oat(s) 1457 

oath 1386 

odd 1339 

of 1258 

off 1258 

offal 1270 

often 1 271 

old 109 

on 1274 

once 1434 

one 1428 

only 1433 

ooze 3058 

open I 341 

or 1418 



orchard laio 
ordeal 1213 
ore 2038 
other 2048 
otter 1314 
ought 1453 
our 1944 
out 1987 
outlaw 33j^ 
ouzel 3050 
oven i3(So 
over 1359 
owe 1453 
owl 1950 
own 1454 
ox 1301 



P 

pail 350 
pan 337 
park 30 
path 133 
pea(oock) 1687 
pear 664 
pease 785 
pebble 41 3 
pen 843 
penny 834 
pike 1 9 10 
pile 1848 
pile 1849 
pillow 1 107 
pine vh 1897 
pine (tree) 

1898 
pipe 1937 
pit 1166 
pitch 581 
pith 460 
play 899 
plight 598 
plot 1333 
plough 31 3 I 
pluck 1058 
plum 1979 
poke 1385 
pole 1385 
pool 3047 
pope 1478 

poppy 1345 

port I3I3 
post 1356 
pound vb 1045 
pound 1974 
pox 1399 
pretty 383 
prick 583 
prickle 583 
pride 3037 



priest 1784 
proud 1994 
prove 3080 
provost 1437 
pull 981 
punt 103a 

q 

quake 398 
qualm 98 
quean 831 
queen 1633 
quell 730 
quench 814 
quick 579 
quicksand 867 
qnid 633 
quoth 131 



race 1394 
rafter 181 

»g352 
raid 1466 
rain 908 
raise 906 
rake 376 



2p 
365 



ram 
ran 30 
rank 189 
ransack 337, 

281 
rash 140 
rather 117 
raven 163 
raw 1 701 
reach 1527 
reach 1528 
read 1583 
TeaAprt 1594 
reap 958 
rear 1481 
reck 870 
reckon 861 
red 1743 
rede 1583 
reed 1833 
reek 1641 
reel 1773 
reeve 1633 
rend 848 
rennet 836 
rest 787 
retch 1528 
rhyme 1900 
rib 646 
rick 1725 
rid 955 



INDEX TO FIRST WORD-LIST. 



ridden 627 
riddle 1584 
ride 1934 
ridge 1158 
right 88a 
righteous 1858 
rim 558 
rime 1901 
rind 546 

ring 5", 3 
ripe 193^ 
rise 1856 
riien 46c 
road 1466 
roer 1371 
rook ia88 
rod 3130 
rode 1465 
roe 1306 
1W1355 
rood 2130 
roof 2078 
rook 2100 
room X977 
rooft 2062 
root 2123 
root rb 2125 
rope 1474 
roie «6 1 248 
roie prt 1395 
rot 1315 
roagh 1982 
row «6 1403 
row v6 2067 
rudder 2049 
ruddy 1070 
rue 1790 
run in/ 688 
run pic 969 
rung 1004 
ruBh 473 
ru8t 99X 
ruth 1807 
rye 1157 



Mck 305 
Md388 
■addle 389 

■»^d 355 
■ail 904 

■aith 894 

■ake 280 

■ale 65 

■allow adj 85 

■allow #6 103 

■alt 107 

■aWe 92 

same 258 



■and 247 
■ang 206 
lank 192 

■•P405 
■at 363 

Saturday 364 

■awjvr^ 312 

■aw 328 

■aw 329 

■•7893 
■ay^894 

■cab 424 

■oale(S8 

■oale^i) 69 

■ehool 2044 

■oorch I3i8 

■core 1178 

■crape 407 

■cuxf 968 

■cythe 1851 

sea 1479 

seal 753 

seam 17x8 

■ear 1690 

seat 1540 

■edge 925 

■ee 1756 

■eed 1586 

■eek 1642 

■eem 1636 

■een 1628 

■eeihe 1778 

■eldom 7<i6 

■elf 738 

■ell 722 

■end 849 

■ent 850 

set 043 
■ettle 930 
■even 804 
■ew 1792 
shabby 424 
shackle 286 
shade 390 
shadow 390 
shaft 182 

»W 353 
shake 285 

shaU67 

shame 259 

shank 193 

shape 406 

shard C2 

(plough^^- 

shm 5 

share 6 

■harp 59 

■have 104 

■haw 333 

she 1755 



sheaf 1 715 
shear 653 
sheath 1480 
shed tb 1169 
shed v6 1468 
sheen 1639 
sheep 1596 
sheer 1482 
sheet 1 661 
shelf 740 
shell 725 
shepherd 1597 
shwd 52 
sheriff 1623, 

1 841 
shew 1703 
shield 767 
shift 498 
shilling 440 
shimmer 560 
shin 527 
shine 1890 
ship 642 
shire 1840 
shirt 1098 

shoal 1322 

shod 2 131 
shoe 2033 
shone 1436 
shook 3103 
shoot 1839 
shorn 11 79 
short I3II 
shot 1319 
should 1240 
shoulder 087 
shove X2O3, 

1967 
shovel 1202 
show 1703 
shower 1947 
shrank 194 
shred 1744 
shrew 1704 
shrift 499 
shrike 1907 
shrine 1891 
shrink 505 
shrive 1876 
shroud 1997 
shrunk 1000 
shun 1018 
shut 1 162 
shuttle 1161 
sick 1818 
sickle 569 
side 1925 
sieve 400 
sift 490 
■igh 1905 



■ight 593 

nlk 445 
■ill I109 

■illy X502 

■ilver 739 

■in 1 138 

■inoe 461 

■inew 526 

■Jngr 515 

■inge 821 
■ink 503 
sister 788 
■it 621 
■ix 880 
■ixth 881 
skate 365 
■kill 433 
skin 531 
sky 2004 
slsdc 282 
slain 330 
slake 283 
slaughter 317 
slaver 163 

■lay 331 

sleage(ham- 
mer) 926 
•leep 1595 
sleeve 1635 
sleight 1659 
slept I59|0 
slew 3116 
sb'd 630 
slide 1926 
slight 884 
slime 1903 
sling 826 
slink 504 
slippery 641 
slit 609 
Bloe 1357 
sloth 1505 
slough 21 15 
slow 1408 
slumber 1049 
slung 1007 
•ly i<554 
■mall 66 
■mart 708 
■mear 652 
■mell 724 
■melt 760 
■mirk 696 
■mite 19 16 
■mith455 
■mithy 462 
smitten 6x0 
■mock 1 291 
■moke 1280 
■mote X459 



399 

smooth 2051 
snail 332 
snake 284 
■xuure 4 
■neak 1906 
■neese X782 
■nipe 191 5 
■not X318 
■now X409 
■o 1358 
■oak 1279 
■oap 1475 
■ock 1290 
■odden 1334 
■oft 2081 
■old 112 
■ome 1047 
■on X0X7 
■ong205 
■oon 2082 
■oot 2x24 
■00th 2050 
■op 1348 

■o" 1373 
■orrow 1206 

•orry 1374 

■ought 21 13 

■onl 1407 

■ound s5 1036 

woundiMdj 1037 

■our 1946 

■outh 1952 

■outhem 1953 

■ow j6 1061 

90W vb 1406 

■liade 391 

■pake 288 

wpui prt 328 

■pan 239 

■par 9 

■pare 10 

■park 4X 

sparrow 24 

■pat »543 
speak 863 

spear 654 

speck 872 

sped X676 

speech 1573 

speed X669 

spell 736 

spelt 762 

spend 851 

spew X869 

spm449 

■pin 532 

spindle 528 

spire 1842 

spit $b 6x I 

spit vb 622 



4CO 



HISTORY OF ENGLISH SOUNDS. 



spoke prt 388 
spoke tb 1449 
spoken laSi 
spoon 20S3 
spot 1338 
sprang 208 
spread 1 547 
spread prt 

1549 
spring 518 

sprang loio 

spun 1033 ■ 

spur 1 180 

spurn 970 

staff 165 

stairs 1533 

stake 387 

stalk 99 

staU78 

stalwart 69 

stammer 360 

stamp 368 

stand 348 

stank 195 

staple 408 

star 667 

stare 8 

stark 40 

starling 7 

starve 685 

staves 166 

stead 947 

steady 948 

steak 915 

steal 713 

steam 17 19 

steed 1668 

steel 1614 

steep 1740 

steeple 1680 

steer 1607 

stem sh 859 

stem vh 858 

stench 811 

step 959 

step- 1835 

stem adj 691 

stem «& 1773 

steward 191 3 

stick 571 

stick 587 

stiff 1877 

stile 600 

still 441 

Bting 517 

stink 505 

stint 1 1 41 

stir Z076 

stirrup 1914 

stitch 570 



stithy 777 
stock 1393 
stolen 1333 
stone 1437 
stood 2133 
stool 30^3 

stoop 3003 

stop 1349 
stork 1303 
storm 1200 
stove 1264 
stow 3069 
strand 349 
strap 1350 
straw 1705 
stream 1720 
street 1578 
strength 835 
stretch 871 
strew 796 
stricken 572 
stride 1927 
strike 1908 
string 823 
strip 168 1 
strode 1469 
stroke 1448 
strong 207 
Btrop 1350 
stubble 1 1 75 
stun 1019 
stung Z009 
stunted 1031 
stye 599 
such 748 
suck 1980 
summer 1048 
sun ro22 
sunder 1038 
sung 1006 
sunk 999 
sup 2001 
swaddle 776 
swain 911 
swallow ib 86 
swallow 1^6754 
swam 266 
Bwan 218 
sward 51 
Bware 3040 
swarm 36 
swarthy 46 
swathe 775 
sway 917 
swear 651 
swrat 1 541 
sweep 1476 
sweet 1660 
swell 723 
sv. erve 684 



swift 497 
swill 432 
swim 562 
swine 1889 
swing 516 
swollen 1228 
swoon 21 17 
sword 710 
swore 2040 
sworn 1177 
swum 1054 
swung ioc8 



tail 345 
take 299 

tale 74 

tallow 105 

tan 236 

tap 420 

tape 419 

taper 411 

tar 661 

tarry 702 

tart 48 

tatter 382 

taught 1 531 

teach 1529 

team 1721 

tear vb 662 

tear «6 1693 

tease 1495 

teem 1037 

-teen 1635 

teeth 1617 

tell 732 

ten 1634 

tetter 940 

than 1276 

thane 910 

thank 191 

that 362 

thatch 279 

thaw 1404 

the 649 

thee 1601 

theft 1626 

their 901 

them 914 

then 1275 

thence 217 

there 1555 

these 467 

thews 1702 

they 892 

thick 586 

thief 181 1 

thigh 1820 

thimble 2023 



thin 1137 
thine 1888 
thing 514 
think 1 1 30 
third 636 
thirst 1082 
thirteen 1828 
thirty 608 
this 466 
thistle 478 
thither 629 
thole 1 22 1 
. thong 204 
thorn 1 198 
thorough 973 
those 1396 
thou 1939 
though 1733 
thought 21 1 2 
thousand i960 
thrall i486 
thrash 679 
thread 1585 
threat(en) 

1739 
three 1754 

thresh 679 

threshold 680 

threw 1 79 1 

thrice 1868 

thrift 495 

thrill 1075 

thrive 1875 

throat 1317 

throe 1356 

throng 203 

through 973 

throw 1405 

thrush 1 1 20 

thrust 2017 

thwart 707 

thumb 1978 

thunder 1016 

Thursday! 945 

thus 988 

tick 580 

tickle 614 

tide 193 1 

tidings 1853 

tie 1656 

tight 883 

til 435 
tile 603 
till 436 
timber 567 
time 1904 
tin 529 
tinder 1145 
tippet 42 1 
tire 663 



tithe 1827 
titmouse 1397 
to 2035 
toad 1^7 I 
toe 1365 
together 397 
token 145 1 
told 115 
toll 1230 
tongs 213 
tongue 1 01 3 
too 2034 
took 2107 
tool 2046 
tooth 205 a 

top 1354 
torn 1187 
tough 21 1 1 
tow 1257 
towards 55 
town 1970 
trap 422 
tread 952 
tree 1761 
trim 1149 
trodden 1336 
troth 1809 
trough 1310 
trow 1804 
true 1803 
trundle 1146 
trust 907 
truth 1808 
Tuesday 1870 
tug 1308 
tumble 1056 
tun 1028 
turf 967 
turn 972 
turtle(dove) 

976 
tusk 990 
twain 1657 
twelfth 743 
twelve 741 
twenty 845 
twice 1 87 1 
twig 604 
twin 537 
twine 1895 
twinkle 510 
twist 484 
twit 19 1 7 
twitch 591 
two 1360 

U 

udder 1995 
ugly 1064 



INDEX TO FIRST WORD-LIST. 



-101 



nnoouih 1956 
under 1033 
untoward 2036 
np 2000 
upon 1073 
n«i957 
utter 1989 



vat 369 
▼ane 221 
vixen 1155 



wade 392 
wag 334 
wail 896 
wain 335 
waist 145 
wake 289 
waken 290 
walk xoo 
wall 79 
wallow 87 
walnut 70 
wan 230 
wand 250 
wander 251 
wane 219 
want 241 
wanton 1309 
ward 53 
-ward 54 
ware 11 
warm 37 
warn 12 
warp 60, I 
wart 47 
wai 128 
waah 142 
wasp 155 
waate 1021 
watch 306 
water 366 



wattle 367 
wave v6 167 
wave «6 1573 
waver 168 
wax v& 309 
wax sb 310 
way 895 
we 1602 
weak 916 
weal 714 
wean 830 
weapon 1598 
wear 657 
weaiT 1608 
weatner 949 
weave 805 
weasel 783 
web 961 
wed 956 
wedge 927 
We£iesday 

3133 
weed 1833 

week 574 

ween 1630 

weep 1682 

weevil 491 

weft 809 

welkin 1233 

well adv 7f3 

well tb 727 

Welsh 715 

wen 837 

wend 852 

weight 596 

weiid 1102 

were 1556 

west 789 

wet 1579 

wether 778 

wey 1533 

wich(elni) 573 

wiok864 

wicked 588 

wide 1928 

widow 631 

wield 768 

wier 655 



wife 1878 
wight 595 
wild 450 
will 442 
willow 755 
win J33 
winch 506 
wind 549 
windlass 550 
window 551 
wine 1892 
wins 823 
wink 507 
winnow 552 
winter 539 
wipe 1935 
wire 1843 
wisdom 1866 
vdse adj 1857 
wise Mb 1859 
wish 2016 
wit 623 
(to) wit 612 
witch 588 
with 456 
without 1990 
withy 457 
wizened 469 
whale 71 
what 368 
wheat 1542 
wheel 177^ 
wheeae 1566 
whelk 715 
whelp 772 
when 1277 
whenoe 220 
where 1557 
whet 944 
whether 119 
whey 1574 
which 749 
while tb 1844 
whileocfr 1845 
whilst 1845 
whine 1893 
whirl 428 
whisper 487 



whistle 479 
whit 596 
white 1919 
whither 632 
who 1360 *- 
whole 1379 
whom 1444 
whoop 2143 
whore 2039 
why 2005 
woad 1469 
woe 1359 
woke 2104 
wolf 983 
woman 1879 
womb 271 
women 1880 
won 1024 
wonder 1040 
wont 1020 
woo 2118 
wood 107 1 
woodruff 2077 
woof 2066 
wool 978 
word 1 215 
work «6 697 
work vh 1089 
world I] 81 
worm 1087 
wormwood 

656, 2138 
worry 1093 
worse 1080 
worship 677 
worst 1077 
wort 1099 
worth 676 
wot 1460 
would 1 24 1 
wound ib 1039 
wound ptc 

1041 
woven 1265 
wrath 1 49 1 
wreak 866 
wreath 1490 
wreck 865 



wren 1522 
wrench 812 
wrest 1500 
wrestle 1501 
wretch 873 
Wright 1091 
wring 519 
wrinkle 508 
wrist 480 
write 191 8 
writhe 1852 
written 613 
wrong 200 
wrote 1 46 1 
wroth 1389 
wrought 1 205 
wrung 101 1 



yard 56 
yard 711 
yam 33 
yarrow 26 
yawn 833 
ye 1604 
yea 1552 
year 1559 
yearn 692 
yeast 792 
yell 731 
yellow 736 
yelp 773 
yes 468 
yesterday 794 
yet 937 
yew 1786 
yield 770 
yoke 1284 
yolk 751 
yon 832 
yore 1692 
you 1787 
young 827 
your 1788 
youth 1826 
yule 1775 



Dd 



402 HISTOEI OF ENGLISH SOUNDS. 



TABLES. 
I. SOUND-CHANGE. 

(Referenoes to §§.) 

Aa. INTERNAL ISOLATIVE. 
voioe to breath, through whisper, esp. in stops (50). also in vowels. 

Vowels. 

short widened, long narrowed (53). 

high become unsyllabio (54). 

place, short lowered, long raised (55). back to front, gen. through mixed (57). 
front to mixed (59). 

rounding, back to rounding, front to unrounding (60). abnormal (6a). 

diphthonging of long vowels (63). smoothing of diphthongs (70). short 
diphthongs (73). glide to cons. (74). 

loss (75). 

oontraotlon (78). 

Consonants. 

form, toeakeninsf* : stop to open, esp. when voiced (79). nasal to c^n (81). 
stop to side (82). side to open (83). open to vowel (84). open breath to breath 
glide (87). BtrengtheningB : open to stop (88). trilling (89). 

place, back to throat (90). front to back (9 a). forward to back (93). 
forward to front (94). lip to lip-teeth (95). forward to lip (96). other changes 
of forward (98). inversion (100). rounding, esp. of back open (102). I (104). 
cleaving (105). smoothing (106). loss (107). addition (108). 

Quantity. 

unstrest sounds shortened; short vowel + long cons, and vice* versa; oonss. 
shortened (no), high vowels shortened (11 a), length-shifting in diphthongs 
(113). cons. -influence (114). different languages (116). infl. on other changes 

("7). 

Foroe. 
force-shifting in diphthongs (118). alternation in stress (119). want of stress 
( 1 20). free stress (121). stress-shifting (122). influence on other changes (124). 

Intonation. 

word-intonation (127). sentence-intonation (128). connection with stress (129). 
to glottal stop (132). infl. on other changes (133). 

Transposition, 
generally isolated (135). 



TABLES : SOUND-CHANGE. 403 



Ab. INTERNAL COMBINATIVE. 

forwardji, backwards; partia], complete (136). 

breath and voioe (137). aspiration (140}. 

vowel harmony (141^ 

ftx)nt-modifieation (142). mutation (145). diphthongic motation (147}. 

baok-inflnenoe (150). 

rounding (151). 

nasalising (153). modifies vowel-formation (156). 

parasite- vowels (159). 

other influenoes (i63\ point (164). height and narroifmess of vowels (165). 
front vowel developed before « (165). s opens stop (166). change of place (167). 
kwXop (168). development of i/eto (169). #; to/(i72). 



B. ACOUSTIC. 

isolative and combinative (173). vowel-pairs (174). striving after audibility 
(176). some changes partly acoustic, partly organic (176). 

C. EXTERNAL CHANGES. 

formal analogy (178). logical analogy (179). popular etymology (180). com- 
plete and parlisl influence (181). blendings (181). differentiation (182). re- 
tardation of oiganio change (183). 

D. GENERAL PRINCIPLES. 

principles of economy : (a) dropping of superfluous sounds ; (6) ease of tran- 
pition (185). economy of exertion doubtful (186). fluctuation (187). 



D il 2 



404 



HISTOEY OF ENGLISH SOUNDS. 



n. FORMS OF LETTERS. 



Capftali. 


HinuKulM. 


Bkck Letter. 


CuxBiw. 


A 


a 


a 


a 


AE 


»S 


m 


a 


B 


b 


* 


h 


C 


c 


t 





D 


bdtS 


n 


d 


E 


e 


e 


e 


F 


F^ 


t 


f 


6 


2g 


a 


9 


H 


h 


ft 


k 


I 


»ij 


9 9 




K 


k 


k 


k 


L 


1 


I 


I 


M 


m 


m 


m 


N 


n 


tt 


n 


O 








o 


P 


P 


V 


p 


Q 


q 


q 


\ 

9' 


B 


pr 


r 


1 
r 


8 


s,f 


s 


s 


T 


ct 


t 


t 


V 


V u 


t)U 


V K 


vv 


w 


to 


10 


X 


z 


X 


X 


T 


yy 


S 


V 


z 


»9 


J 


z 



TABLES : MODERN ENGLISH VOWELS. 



407 



VI MODERN ENGLISH VOWELS. 



1MB 


fMn 


■Bill 


thMn 


LB 


. \ 


ffUMI 


», a 


» 


» 


» 


* i 


path 


e, a 


M 


aee 


aa 


• 

1 


wU 


• 
% 


• 
% 


ft 


• 
ft 


e 


0nd 








6 


6 


n 


ton 


« 


A 


« 


« 





ox 





l> 





9 


ft 


name 


MB,aa 


«0 


ee 


0l,6i 


1 


wine 


ei 


9i 


•i 


6i,ai 


S 


ffreen 


ii 


U 


66, U { 


<i 


i 


deal 


00 


ae 


a 


hoU00 


oa 


•a 


6a 


60, »Q 


s 


moon 


au 


na 


XOL 


IfW, «W 


i 


stone 


00 


00 


00 


oa, aa 


ai 


day 
ihey 


ai, 00 


ei, 00 \ 
00 1 


66 


«i,6i 


d 


hoU 


oi, ui 


oi, Ai 


d,6i 


oi,6i 


sa 


9ttW 


an, 00 


00 


00 


00 


ea(S) 


new 


yy(n), iu 
0a 


77* »« ) 
en, ia 1 


Jan 


JflW,j w 


Ga 


grow 1 
know 1 


on 


oity 00 


00 


00, ao 



CONTRACTIONS. 



^fo « MSne. 

JEfciR - iSnfrio's Homilies. 

JEpr - M\>e]i9d. 

AFrinan » Anglo-Friaian. 

Aldhgl « Aldhehn Glosiafy. 

AllP B Morrises Alliterative Pocsob. 

Amer. ■■ American. 

AR » Ancren Biwle. 

Ar. a Arian, Arabic. 

And. B Audelay. 

Aj. B Ayenbite. 

Bob ■■ BacbanazL 

Bert. B Bertiaiy. 

Bff s Bulgarian. 

BH -» Bede's Hirtory (OET). 

BIH » Blickling HomiUes. 

Bll - Bollokar. 

Bo* = Boulogne. 

Boetb. B BoetbiuB. 

Bt » Butler. 

Bulg. — Bulgarian. 

GH B Csedmon's Hymn. 

ob(art.) « obarter. 

Cb s Cbauoer. 

Gbr(on.) « Cbronide. 

cj a oonjunction. 

Gk » Gbeke. 

GM B GuTBor Mundi. 

op ■■ comparOp oomparatiTe. 

(^ B GoipuB gloisary, Gooper. 

Dan. B Daniflb. 

Du. B Durbam GoBpeU. 

e- s Early. 

£. — Engliab. 

E- s East. 

EO s Expert Ortbograpbirt. 

£p. B Epinal gloBsary. 

E(r)f: - Erfurt. 

f- « First. 
Fk B Franklin. 
F(r) « Flrenob. 

GaeL ib Gaelic 

GE «■ GenesiB and EzoduB. 

Gk s Greek. 



-gl a GloBBary. 

G(m) B G^erman. 

Gmc B Germanio. 

Grotb. as Gotbic. 

Grein ■■ Groin's Bibliotbek der angel- 

sacbBiscben Poesie. 
Gu>L a Godwin*B Gu))lac 

Harl. B Harleian MS. 

Hom. B Morris's Old-Englisb Homilies. 

Ht -Hart. 

Hung. B Hungarian. 

Hy b Havelok. 

HVg B Welsb Hymn to tbe Yiigin. 

loel. B (Modem) Icelandic. 
ItaL B Italian. 

Jn B Jones. 

Jul. B St. Juliana. 

Eatb. B St. Katbarine. 
KS B Kentiab Sermons. 
Kt B Kentisb. 

1- B Late. 

L« B livizig. 

Lay. B Layamon. 

Ld B Laud (Peterborougb) Gbronicle, 

Lediard. 
Leecbd. b Gookayne*s Leecbdoms. 
Lei. B Leiden. 
Litb. B litbuanian. 
L(a)t B Latin. 
LV B Liber Vltae (OET). 

M- - Middle. 

Marg. B St. Margaret. 

Merc B Mercian. 

Mg B Miege. 

MH B Metrical Homilies. 

MHG B Middle Higb Gennan. 

Ml B Midland. 

Mn- B Modem. 

Mod. B Modem. 

Mtt B Mattbew. 

N-BKortb. 

Kortb. B NortbnmbriaDy Nortbem. 

Norw. B Norwegian. 



TABLES: ENGUSH VOWELS. 



405 



m. ENGLISH VOWELS. 



OB 


ME 


LE 


ma-UTi 


man 


man 


net 


sat 


sset 


heard 


hard 


haad 




name 


neim 


witan 


witen 


wit 


belpan 


helpen 


help 


heofon 


hevene 


heTn 


stdan 


stolen 


Btijl 


^nde 


ende 


end 


m^ 


m6te 


mijt 


■nnii 


Bune 


ivn 


synn 


sinne 


Bin 


ooca 


OExe 


dkB 


open 


dpen 


oupn 


BtSn 


Bt(n 


Btoun 


d£l 


dfl 


dijl 


dr6am 


dr^m 


drijm 


win 


win 


wain 


gr6ne 


gr6ne 


grijn 


dSop 


d6p 


dijp 


hOB 


hUB 


haiiB 


mod 


mod 


mnwd 


lyr 


fir 


&for 



i>d3 



4o6 



HISTORY OP ENGLISH SOUNDS. 



IV. OLD-ENGLISH DIALECTS. 



Omc 


eWS 


IWB 


eKt 


IKI 


Were. 


North. 


( 


monii) a 


a 


o, a 


a 





1 

o 


f 




heard 


ea 


ea 


ea 


ea 


ea, a 


» i 




6aU,a 


ea 


ea, a 


ea 


a 


a 






geseali 


ea 


ea 


ea 


» 


se 




\ 


geaf 


ea 


» 


8B 


SB 


m 


6 


weorc 


eo 


eo 


eo 


e 


e 


a-i 


ierfe 


y 


« 


« 


« 


% 


• 

1 


bierhtQ 


y 


• 

1 


i 


i 


i 


g« 


gieUan 


y 


e 


e 


e 


e 


ka-i 


dele 


y 


« 


« 


e 


9 


u-i 


synn 


y 


y 


e 


y 


y 


o-i 


ele 


e 


<B 


e 


e 


€B 


£ 


dsd 


£ 


8 


8 


8 


8 


j* 


gSar 


8a 


8 


6 


8 


8 


an-i 


hieran 


y 


8 


8 


8 


8 


eu-i 


geiiene 


y 


8 


8 


8 


8 


aug 


Sage 


8a 


6a 


8a 


8 


8 


eng 


flSogan 


So 


8o 


So 


8 


8 


n-i 


^ 


y 


y 


8 


f 


f 


5-i 


gr6ne 


8 


ol 


8 


& 


m 



V. MIDDLE-ENGLISH DIALECTS. 



OE 


Bth 


EMI 


Kt 


Cb 


o,a 


mon 


a 


a 


a 


» 


^ 


a 


a,e 


a 


ea 


h^ 


a 


a 


a 


ft,ea 


§ld 


ft 


ya 


§ 


eo 


eor]ie 


eo, e 


ye 


e 


y 


sttnne 


i 


e 


i 


ft 


Btgn 


S 


§ 


9 


8a 


dgd 


i 


ya 


i 


80 


d8oTeI 


80,8 


ye 


8 


y 


vUr 


i 


8 


i 



CONTBACTIONS. 



409 



* Ormulum. 
Q. » Old. 

OE « Old EngliBh (Aiifflo-S*xoii). 
0£T - Oldest E^liih Texts. 
OH6 « Old High Gemuui. 

01 « Old loelandio. 

ON - Owl and Nightingale. 
Or. » Orodoi. 

P. - Price. 

Paat. - Paatoral Care. 

PC » Prick of Conioience. 

PCbr B Parker Chronicle. 

Pg » Palamve. 

pd * pluxal. 

Port. ■> Portngaeee. 

PPl - Piers Ploughman. 

pm -■ pronoun. 

Prompt. ■■ Promptoriom Panrolomm. 

prp s preposition. 

prt *B preterite. 

Ps -» MetriMl Psalter. 

pto ■■ partidple. 



RBC « Bobert of Bninne*s Chronicle. 
BG(1) - Robert of Gloucester. 
Bit. - Durham Bitual. 
Bn. -i Bushworth gloss (Matthew). 
Buss. ■■ Bnssian. 

s- * Second. 



8- « South. 

sb *B substantive. 

Sb - Salesbury. 

Scand. ■■ Scandinavian. 

Soint. » liber Scnntillamm. 

sff * singular. 

Sh » Sheridan. 

Sk- Sanskrit. 

Sm * Smith. 

St- « Standufd. 

Sth - Southern. 

Snff. « Suffolk. 

Sw(ed.) » Swedish. 

Td » Tindal. 

th. - Third. 

TM - Townley MTsteries. 

Tr - Transition. 

vb — verb. 

Verc. - VeroeUi MS. 

vg — Vulgar. 

yp - Vespasian Psalter. 

VS « Visible Speech. 

W. - Wallis. 
W- - West. 
Wgl - Wright's 
Wiol. - Widif. 
Wk - Wilkins. 
WS - West-^azon. 



THE SND. 



Works by HENRY SWEET, M.A. 



AN ANGLO-SAXON PRIMER, with Grammar, Notes, and 
Glossary. JSecand Edition, 28, 6d, 

AN ANGZO-^SAXON READER. In Prose and Verse. With 
Grammatical Introduction, Notes, and Glossary. Fowrth Edition y 
Revised and Enlarged, 8«. td, 

A SECOND ANGLO-SAXON READER. Archaic and Dia- 
lectal. 4«. 6d, 

OLD ENGLISH READING PRIMERS:— 

I. Selected Homilies of ^Ifric. is. 6(2. 
II. Extracts from Alfred's Orosins. is. 6d, 

' In thoi publishing these Early English Texts at a low price and in a 
convenient form, Mr. Sweet confers a real boon on the students of the 
language, and we hope he wiU fulfil his promise of continuing the series 
till he has issued all the important works of old Eiiglish literature.* — 
Saturday Retiew. 

FIRST MIDDLE ENGLISH PRIMER, with Grammar and 
Glossary. 2$, 

SECOND MIDDLE ENGLISH PRIMER. Extracts from 
Chaucer, with Grammar and Glossary. 28, 

* The most important feature of this little book is the oopions information 
which it gives respecting the pronunciation of English in Chaucer s time.* — 
AiheiKBum, 

AN ICELANDIC PRIMER, with Grammar, Notes, and Glos- 
sary. 3*. 6d, 

A HANDBOOK OF PHONETICS, including a Popular Exposi- 
tion of the Principles of Spelling Reform. 48. 6d, 

ELEMENTARBUCH DES GESPROCHENEN ENGLISCH. 
' Grammatik, Texte und Glossar. Second Edition, 28. 6d. 



Oxtovb 

AT THE CLARENDON PRESS 

LONDON : HENRY FROWDE 
OxFOKD UNivEBsrrr Pbess Warehocse, Aukh Corxer, E.C. 



May, 1888. 



iEIatentron iPtess^ <!^xfotli. 

A SELECTION OF 

BOOKS 

PUBUSHED FOR THE UNIVERSITY BY 

HENBY FBOWDE, 

AT THE OXFORD UNIVERSITY PRESS WAREHOUSE, 

AMEN CORNER, LONDON. 

ALSO TO BB HAD AT THB 

CLARENDON PRESS DEPOSITORY. OXFORD. 
\Ev€ry hook is bound in cloth, unless otherwise described.'] 



LEXICONS, GHAMMABS, OBIENTAL WOBES, fto. 

Anglo-Saxon. — An Anglo-Saxon Dictionary^ based on the 

MS. Collections of the late Joseph Bosworlh, D.D., Professor of Anglo-Saxon, 
Oxford. Edited and enlamd by Prof. T. N. Toller, M.A. (To be completed 
in four parts.) Parts I-Ilf. A — SAR. 4to. 15/. ^ch. 

Arabic. — A Practical Arabic Grammar. Part I. Compiled 

by A. O. Green, Brigade Major, Royal Engineers, Author of ' Modem Arabic 
Stories * Second Ecution, Enlarged and Revised. Crown Syo. 7/. (m/. 

Chinese.—.^ Handbook of the Chinese Language. By James 

Sammers. 1863. Syo. half bound, i/. 8j. 

A Record of Buddhistic Kingdoms^ by the Chinese Monk 

FA-HI EN. Trandated and annotated by James Legge, M.A., LL.D Crown 
4to. cloth back, lOf. (m/. 

English. — A New English Dictionary^ on Historical Prin- 

Hpla: founded mainly on the materials collected by the Philological Society. 
Edited by James A. H. Murray, LL.D., with the assistance of many Sdiolars 
and men of Science. Part I. A— ANT. Part IL ANT— BATTEN. 
PartllL BATTER— BOZ. Imperial 4to. la/. Meach. 

B 



2 CLARENDON PRESS, OXFORD. 

English. — An Etymological Dictionary of the English Lan- 
guage. By W. W. Skeat. Litt.D. Second Edition, 1884. 4to. a/. 4J. 

Supplement to the First Edition of the above. 4to. aj. 6rf. 

A Concise Etymological Dictionary of the English Lan^ 

guage. By W. W. Skeat Litt.D. Third Edition. 1887. Crown 8vo. fx. 6*/. 

Greek. — A Greek-English Lexicon, by Henry George 

Liddell, D.D., and Robert Scott, D.D. Seventh Edition, Revised and Aug- 
mented throughout. 1883. 4to. il. 16s. 

A Greek-English Lexicon, abridged from Liddell and 

Scott's 4to. edition, chiefly for the use of Schools. Twenty-first Edition. 
1884. Square 1 2mo. js.od. 

A copious Greek-English Vocabulary^ compiled from 



the best authorities. 1850. 34mo. 3/. 

A Practical Introduction to Greek Accentuation, by H. 

W. Chandler, M. A. Second Edition. 1881. 8vo. iox.6</. 

Hebrew. — The Book of Hebrew Roots, by Abu '1-Waltd 

Marw&n ibn Jan&h, otherwise called Rabbt Ydnih. Now first edited, with an 
Appendix, by Ad. Neubauer. 1875. 4to. a/. *js. td. 

A Treatise on the use of the Tenses in Hebrew. By 

S. R. Driver, D.D. Second Edition. 1 881. Extra fcap. 8vo. 7/. 6^. 

Hebrew Accentuation of Psalms, Proverbs, and Job. 

By William Wickes, D.D. 1881. Demy 8vo. fJ. 

A Treatise on the Accentuation of the twenty-one so-called 

Prose Books of the Old Testament. By William Wickes, D.D. 1887. Demy 
8vo. ios,6d. 

Icelandic. — An Icelandic- English Dictionary^ based on the 

MS. collections of ^he late Richard Cleasby. Enlaiged and completed by 
G. Vigfusson. M.A. With an Introduction, and Life of Richard Cleasby. by 
G. Webbe Dasent, D.C.L. 1874. 4to. 3/. 7/. 

A List of English Words the Etymology of which is 

illustrated by comparison with Icelandic. Prepared in the form of an 
Appendix to the above. By W. W. Skeat, Litt.D. 1876. stitched, a/. 

'An Icelandic Primer, with Grammar, Notes, and 



Glossary. By Henry Sweet, M.A. Extra fcap. 8vo. 3J. (id. 

An Icelandic Prose Reader, with Notes, Grammar and 

Glossary, by Dr. Gudbrand Vigfusson and F. York Powell, M.A. 1879. 
Extra fcap. 8vo. los. 6</. 

Latin. — A Latin Dictionary, founded on Andrews' edition 

of Freund*s Latin Dictionary, revised, enlarged, and in great part rewritten 
by Charlton T. Lewis, Ph.D., and Charles Short, LL.D. 1879. 4to. i/. 5^ . 



CLARENDON PRESS, OXFORD. 



Melanesian. — The Melanesian Languages, By R. H. 

Codrington, D.D.,^ of the Melanesian Mission. 8vo. i8j« 

Sanskrit. — A Practical Grammar of tlu Sanskrit Language^ 

arranged with reference to the Classical Languages of Europe, for the use of 
English Students, by Sir^M. M on ier- Williams, M.A. Fourth Edition. 8yo. 

A Sanskrit'English Dictionary^ Etymologically and 



Philologically arranged, with special reference to Greek, Latin, German, Anglo- 
id other cognate Indo»Europei 
Monier- Williams, M.A. 1888. 4to. 4/. 14/. 6a. 



Saxon, English, and other cognate Indo-European Languages. By Sir M. 



Nalopdkhydnam. Story of Nala, an Episode of 

the Mahi-Bhiirata: the Sanskrit text, with a copious Vocabulary, and an 
improved version of Dean Milman*s Translation, by Sir M. Monier- Williams, 
M.A. Second Edition, Revised and Improved. 1879. 8vo. 15/. 

Sahuntald, A Sanskrit Drama, in Seven Acts. Edited 

by Sir M. Monier- Williams, M.A. Second Edition, 1876. 8vo. aix. 

Syriac. — T/tesaurus Syriacus : coUegerunt Quatrem^re, Bern- 
stein, Lorsbach, Amoldi, Agrell, Field, Roediger: edidit R. Payne Smith, 
S.T.P. Fascl-Vl. 1868-83. sm. fol. each, i/. i«. Fasc. VIL iLiis,6d. 

Vol. I, containing Fasc. I-V, sm. fol. 5/. 5/. 

TAe Book ofKalllah and DimnaJu Translated from Arabic 

into Syiiac. Edited by W. Wright. LL.D. 1884. 8vo. au. 



GBEEE CLASSICS, fto. 

Aristophanes: A Complete Concordance to the Comedies 

and Fragments. By Henry Dunbar, M.D. 4to. 1/. is, 

Aristotle: The Politics, with Introductions, Notes, etc., by 

W. L. Newman, M.A., Fellow of Balliol College, Oxford. Vols. I. and 11. 
Medium 8vo. 28^. Just Published, 

Aristotle: The Politics, translated into English, with Intro- 
duction, Marginal Analysis, Notes, and Indices, by B. Jowett, M.A. Medium 
8vo. 2 vols. 2 1 J. 

Catalogus Codicum Graecorum Sinaiticorum, Scripsit V, 

Gardthausen Ltpsiensls. With six pages of Facsimiles. 8vo. limn, 25/. 

Heracliti Ephesii Reliquiae. Recensuit I. Bywater, M.A. 

Appendicis loco additae sunt Diogenis Laertii Vita Heracliti, Particulae Hip- 
pocratei De Diaeta Libri Primi, Epistolae Heracliteae. 1877. 8vo. dr. 

Herculanensium Voluminum Partes II. 1824. 8vo. 10s. 

B 2 



CLARENDON PRESS, OXFORD. 



Fragmenta Herculanensia, A Descriptive Catalogue of the 

Oxford copies of the Hercnlanean Rolls, together with the texts of several 
papyri, accompanied by facsimiles. Edited by Walter Scott, M.A., Fellow 
of Merton College, Oxford. Royal 8vo. cloth, aix. 

Homer: A Complete Concordance Jto the Odyssey and 

H3rmns of Homer ; to which is added a Concordance to the Parallel Passages 
in the Iliad, Odyssey, and Hymns. By Henry Dunbar, M.D. 1880. 4to. i/. is^ 

Scholia Graeca in Iliadent. Edited by Professor W. 

Dindorf, after a new collation of the Venetian MSS. by D. B. Monro, M^.^ 
ProTOst of Oriel Coll^;e. 4 vols. 8vo. 2/. lor . Vols. V and VI. In the Press. 



- Scholia Graeca in Odysseam. Edidit Guil. Dindorfius. 

Tomill. 1855. 8vo. i^.6d. 

Plato : Apology, with a revised Text and English Notes, and 

a Digest of Platonic Idioms, by James Riddell, M.A. 1878. 8vo. Zs. 6d, 

Philebus, with a revised Text and English Notes, by 

Edward Poste, M.A. i860. 8vo. 71. 6</. 

Sophistes and Poliiicus, with a revised Text and English 

Notes, by L. Campbell, M.A. 1867. 8yo. i8j. 

Theaeietus, with a revised Text and English Notes. 



by L. Campbell, M.A. Second Edition. 8yo. lor. 6d. 

- The Dialogues, translated into English, with Analyses 

and Introductions, by B. Jowett, M.A. A new Edition in 5 volumes, medium- 
8vo. 1875. 3/. loj. 

The Republic, translated into English, with an Analysis 



and Introduction, by B. Jowett, M.A. Medium 8vo. I2j. 6d. 

Thucydides : Translated into English, with Introduction, 

Marginal Analysis, Notes, and Indices. By B. Jowett, M.A. a vols. x88i. 
Medium 8vo. i/. lax. 



THE HOLY SCBIPTUBES, fto. 

Studia Biblica. — Essays in Biblical Archaeology and Criti- 
cism, and kindred subjects. By Members of the University of Oxford. Svo. 
I Of. 6</. 

English.— 7'*^f Holy Bible in the earliest English Versions, 

made from the Latin Vulgate by John Wycliffe and his followers : edited by 
the Rev. J. Forahalland Sir F. Madden. 4 vols. 1850. Royal 4to. 3/. 3/. 



CLARENDON PRESS, OXFORD. 



[Also reprinted from the above* with Introduotion and Glossary 

by W. W. Skeat, Iiitt. D. 

— The Books of Job, Psalms, Proverbs, Ecclesiasies, and the 

Song of SolcMCH : according to the Wycliffite Version made by Nicholas 
de Hereford, abont A.D. 1381, and Revised by John Purvey, about A.D. 1388. 
Extra fcap. 8vo. 3/. 6d. 

— The New Testament in English, according to the Version 

by John Wyclifie, abont A.D. 1380, and Revised by John Purvey, about a.d. 
1388. Extra fcap. 8vo. 6/.] 



English. — The Holy Bible: an exact reprint, page for page, 

of the Authorised Version published in the year 161 1. Demy 4to. half bound, 
i/. I/. 

The Psalter, or Psalms of David, and certain Canticles, 

with a Translation and Exposition in English, by Richard RoUe of Hampole. 
Edited by H. R. Bnunley, M.A., Fellow of S. M. Magdalen College, Oxford. 
With an Introduction and Glossary. Demy 8vo. i/. u. 

Lectures on the Book of Job. Delivered in Westminster 

Abbey by the Very Rev. Geoige Granville Bradley, D.D., Dean of West- 
minster. Crown 8vo. 7j. 6</. 

Lectures on Ecclesiastes. By the same Author. Crown 



8vo. 4r. Cm/. 

Gothic. — The Gospel of St. Mark in Gothic, according to 

the translation made by Wulfila in the Fourth Century. Edited with a 
Grammatical Introduction and Glossarial Index by W. W. Skeat, Litt D. 
Extra fcap. 8vo. 47. 

Greek. — Vetus Testamentum ex Versione Septuaginta Inter- 

pretum secundum exeniplar VaticanumRomae editum. Accedit potior varietas 
Codids Alexandrini. Tomi III. Editio Altera. i8mo. i8j. The volumes 
may be had separately, price 6s. each. 

Origenis Hexc^lorum quae supersunt; sive, Veterum 

Interpretum Graecorum in totum Vetus Testamentum Fragmenta. Edidit 
Fridericus Field, A.M. a vols. 1875. 4^o* 5^* h'* 

The Book of Wisdom: the Greek Text, the Latin 

Vulgate, and the Auth'brised English Version ; with an Introduction, Critical 
Apparatus, and a Commentary. By William J.Deane.M.A. Small4to. 12s. 6d, 

Novum Testamentum Graece. Antiquissimorum Codicum 

Textns in online parallelo dispositi. Accedit collalioCodicisSinaitici. Edidit 
£. H. Hansell. S.T.B. Tomi III. 1864. 8vo. 2^1. 

Novum Testamentum Graece. Accedunt parallela S. 

Scripturae loca, etc. Edidit Carolus Lloyd, S.T.P.R. i8mo. y. 
On writing paper » with wide margin, los. 



CLARENDON PRESS, OXFORD. 



Greek. — Novum Tesiamentum Graece }uxtSi Exemplar Mi Ilia- 

nnm. i8mo. 2s. 6d. On writing paper, with wide margin, 9J. 

Evangelia Sacra Graece. Fcap. 8vo. limp, i j. 6d. 

The Greek Testament, with the Readings adopted by 



the Revisers of the Authorised Version : — 

(i) Pica type, with Marginal References. Demy 8yo. loj. 6</. 

(a) Long Primer type. Fcap. 8yo. 4s. 6rf. 

(3) The same, on writing paper, with wide margin, 15s. 

The Parallel New Testament, Greek and English ; being 



the Aathorised Version, 161 1; the Revised Version, iSSi; and the Greek 
Text followed in the Revised Version. Svo. I3J. 6</. 

The Revised Version is the joint property of the Universities of Oxford and Cambridge. 

Canon Muratorianus : the earliest Catalogue of the 

Books of the New Testament. Edited with Notes and a Facsimile of the 
MS. in the Ambrosian Library at Milan, by S. P. Tregelles, LL.D. 1867. 
4to. 10s, 6d. 

Outlines of Textual Criticism applied to the New Testa- 

nunt. By C. £. Hammond, M.A. Fourth Edition. Extra fcap. 8vo. 3^. 6^. 

Hebrew, etc. — Notes on the Hebrew Text of the Book of 

Genesis. With Two Appendices. By G. J. Spnrrell, M.A. Crown 8vo. 
IQT. 6^. 

The Psalms in Hebrew without points, 1879. Crown 

8vo. Price reduced to aj., in stiff cover. 

A Commentary on the Book of Proverbs, Attributed 

to Abraham Ibn Ezra. Edited from a MS. in the Bodleian Library by 
S. R. Driver, M.A. Crown 8vo. paper covers, 3J. 6^. 

The Book of Tobit. A Chaldee Text, from a unique 

MS. in the Bodleian Library ; with other Rabbinical Texts, English Transla- 
tions, and the Itala. Edited by Ad. Neubauer, M.A. 1878. Crown Svo. 6j. 

Horae Hebraicae et Talmudicae, a J. Lightfoot. A new 

Edition, by R. Gandell, M.A. 4 vols. 1859. Svo. \Uis, 



Latin. — Libri Psalmorum Versio antiqua Latina, cum Para- 

phrasi Anglo-Saxonica. Edidit B. Thorpe, F.A.S. 1835. Svo. ioj. 6</. 

Old-Latin Biblical Texts: No. /. The Gospel according 

to St. Matthew from the St. Germain MS. (g,). Edited with Introduction 
and Appendices by John Wordsworth, D.D. Small 4to., stiff covers, 6j. 

Old-Latin Biblical Texts : No. //. Portions of the Gospels 

according to St. Mark and St. Matthew, from the Bobbio MS. (k), &c. 
Edited by John Wordsworth, D.D., W. Sanday, M. A., D.D., and H. J. White, 
M.A. Small 4to., stiff covers, 21s, 



CLARENDON PRESS, OXFORD. 



Latin — Old-Latin Biblical Texts: No. ILL The Four 

Gospels* from the Munich MS. (q), now nambered Lat. 6224 in the Royal 
Library at Munich. With a Fragment from St. John in the Hof-Bibliothek at 
Vienna (Cod. Lat. 502). Edited^ with the aid of Tischendorfs transcript 
(nnder the direction of the Bishop of Salisbury), by H. J. White, M. A. Small 
4to. stiff covers, 1 2s. 6d. 

Old-French.— Z/^/ Psalmorum Versio antiqua Gallica e 

Cod. MS. in Bibl. Bodleiana adservato, nna cum Versione Metrica aliisque 
Monumentis pervetustis. Nunc primum descripsit et edidit Franciscus Michel, 
Phil. Doc. i860. 8vo. lar. td. 



FATHEBS OF THE CHXJBCH, &o. 

St Athanasius : Historical Writings, according to the Bene- 
dictine Text. With an Introduction by William Bright, D.D. 1881. Crown 
8vo. lox. 6d, 

Orations against the Arians. With an Account of his 



Life by William Bright. D.D. 1873. Crown 8vo. 9/. 

St. Augustine: Select Anti-Pelagian Treatises, and the Acts 

of the Second Council of Orange. With an Introduction by William Bright, 
D.D. Crown 8vo. 9^. 

Canons of the First Four General Councils of Nicaea, Con- 
stantinople, £phesus,and Chalcedon. 1877. Crown 8vo. as. 6tif. 

- Notes on the Canons of the First Four General Councils. 



By William Bright, D.D. 1882. Crown 8vo. 5J. 6d. 

Cyrilli Archiepiscopi Alexandrini in XII Prophetas. Edidit 

p. E. Pusey, A.M. Tomi II. 1868. 8vo. cloth, a/, aj. 

in D. Joannis Evangelium. Accedunt Fragmenta varia 

necnon Tractatus ad Tiberium Diaconum duo. Edidit post Aubertum 
P. E. Pusey, A.M. Tomi III. 187a. 8vo. a/. 5/. 

Cofnmentarii in Lucae Evangelium quae supersunt 

Syriace. E MSS. apud Mus. Britan. edidit R. Payne Smith, A.M. 1858. 
4to. i/. IS, 

Translated by R. Payne Smith, M.A. % vols. 1859. 

8yo. 1 4/. 

Ephraemi Syri, Rabulae Episcopi Edesseni, Balaei, aliorum- 

que Opera Selecta. £ Codd. Syriacis MSS. in Museo Britannico et Bibliotheca 
Bodleiana asservatis primus edidit J. J. Overbeck. 1865. ^^o. i/. is. 



CLARENDON PRESS, OXFORD. 



Eusebius^ Ecclesiastical Historyy according to the text of 

Burton, with an Introdnction by William Bright, D.D. 1881. Crown 8to. 

Irenaeus: The Third Book of St.IrenaeuSy Bishop of Lyons, 

against Heresies. With short Notes and a Glossaiy by H. Deane, BJ>. 
1874. Crown 8vo. $s, 6d. 

Patrum ApostoHcoruniy S. Clementis Romani, S. Ignatii, 

S. Polycarpi, quae supersunt. Edidit Guil. Jacobson, S.T.P.R. Torn! II. 
Fourth Edition, 1863. 8vo. i/. ix. 

Socrates* Ecclesiastical History, according to the Text of 

Hussey, with an Introduction by William Bright, D.D. 1878. Crown 8vo. 



ECCLESIASTICAIi HISTORY, BIOGBAPHT, &o. 

Ancient Liturgy of the Church of England, according to the 

uses of Sarum, York, Hereford, and Bangor, and the Roman Liturgy arranged 
in parallel columns, with preface and notes. By William Maskell, M.A. 
Third Edition. i88a. 8vo. 15X. 

Baedae Historia Ecclesiastica. Edited^ with English Notes, 

by G. H. Moberly,M.A. 1881. Crown 8yo. lor. fid. 

Bright ( W^. Chapters of Early English Church History. 

1878. 8vO. I2X. 

Burnet s History of the Reformation of the Church of England, 

A new Edition. Carefully revised, and the Records collated with the originals, 
by N. Pocock, M.A. 7 vols. 1865. 8yo. Price reduced to x/. loj. 

Councils and Ecclesiastical Documents relating to Great Britain 

and Ireland. Edited, after Spelman and Wilkins, by A. W. liaddan, B.D., 
and W. Stubbs, M.A. Vols. I. and III. 1869-71. Medium 8vo.' each i/. ix. 

Vol. II. Part I. 1873. Medium 8vo. ioj. 6d, 

Vol. II. Part II. 1878. Church of Ireland; Memorials of St. Patrick. 
Stiff covers, 3/. 6d, 

Hamilton {John, Archbishop of St. Andrews), The Catechism 

of. Edited, with Introduction and Glossary, by Thomas Graves Law. With 
a Preface by the Right Hon. W. E. Gladstone. 8vo. 1 2x. 6</. 

Hammond (C -£".). Liturgies, Eastern and Western. Edited, 

with Introduction, Notes, and Liturgical Glossary. 1878. Crown 8vo. loj. dd. 
An Appendix to the above. 1879. Crown 8vo. paper covers, ix. 6</. 



CLARENDON PRESS, OXFORD. 



John, Bishop of Ephesus. The Third Part of his Eccle-^ 

siustUal History. [In Syriac.] Now first edited by William Careton, M.A. 
1853. 4to. i/. lu, 

Translated by R. Payne Smith, M.A. i860. 8vo. 10s. 

Leofric Missal, The, as used in the Cathedral of Exeter 

dnring the Episcopate of its first Bishop, A.D. 1050-1072 ; together with some 
Account of the Red Book of Derby, the Missal of Robert of Jnmi^geSi and a 
few other early MS. Service Books of the English Church. Edited, with In- 
troduction and Notes, by F. £. Warren, B.D. 4to. half morocco, 35/. 

Monumenia Ritualia Ecclesiae Anglicanae. The occasional 

Offices of the Church of England according to the old use of Salisbury, the 
Prymer in £lnglish, and other prayers and forms, with dissertations and notes. 
By William Itlaskell, M.A. &cond Edition. i88a. 3 vols.Svo. 2/. loj. 

Records of the Reformation. The Divorce, 1527-1533. Mostly 

now for the first time printed from MSS. in the British Museum and other 
libraries. Collected and arranged by N. Pocock, M.A. 1870. 2 yoIs. 8vo. 
1/. 1 6 J. 

Shirley ( W. W.). Some Account of the Church in the Apostolic 

Agi, Second Edition^ i874* Ecap. 8vo. 3s. 6d, 

9 

Stubbs [ W.y Registrum Sacrum Anglicoftum. An attempt 

to exhibit the course of Episcopal Succession in England. 1858. Small 4to. 
8j. 6d, 

Warren {F. E.). Liturgy and Ritual of the Celtic Church. 

1881. 8vo. 14J. 

ENGLISH THEOLOQY. 

Bampton Lectures, 1886. The Christian Platonists of Alex- 

cmdricL. By Charles Bigg, D.D. Syo. lor. 6</. 

Butlet^s Works, with an Index to the Analogy. % vols. 1874. 

8yo. IIS. 

Also separately. 

Sermons, 5s. 6d. Analogy of Religion, 5s. 6d. 

Greswelts Harmonia Evangelica. Fifth Edition. 8vo. 9J. 6rf. 
HeurtUys Harmonia Symbolica: Creeds of the Western 

Church. 1858. 8vo. 6/. td. 

Homilies appointed to be read in Churches. Edited by 

J. Griffiths, M.A. 1859. 8vo. 7/. 6d. 

Hooker's Works, with his life by Walton, arranged by John 

Keble, M.A. SeYcnth Edition. Revised by R. W. Church, Af.A., D.C.L., 
Dean of St. Patties* and F. Paget ^ D.D. 3 vols, medium 8vo. 36/. 



lo CLARENDON PRESS, OXFORD. 

«- 

Hooker's Works, the text as arranged by John Keble, M.A. 

2 vols. 1875. 8vo. IIJ. 

Jewers Works. Edited by R. W. Jelf, D.D. 8 vols. 1848. 

8vo. i/. lox. 

Pearson* s Exposition of the Creed. Revised and corrected by 

E. Burton, D.D. Sixth Edition, 1877. 8vo. loj. 6d. 

Waterland's Review of the Doctrine of the Eucharist, with 

a Preface by the late Bishop of London. Crown 8vo. dr. 6^/. 

Works, with Life, by Bp. Van Mildert. A new Edition, 

with copious Indexes. 6 vols. 1S56. 8vo. li. us. 

Wheatly^s Illustration of the Book of Common Prayer. A new 

Edition, 1846. 8vo. fj. 

Wyclif A Catalogue oftJie Original Works of John Wyclif 

by W. W. Shirley, D.D. 1865. 8vo. 3X. 6d, 

Select English Works. By T. Arnold, M.A. 3 vols. 

1869-1871. 8vo. i/. IX. 

Trialogus. With the Supplement now first edited. 

By Gotthard Lechler. 1869. 8vo. 7x. 



HISTOBICAL AND DOCT7MENTABY WOBKS. 

British Barrows, a Record of the Examination of Sepulchral 

Mounds in various parts of England. By William Greenwell, M.A., F.S. A. 
Together with Description of Figures of Skulls, General Remarks on Pre- 
historic Crania, and an Appendix by George Rolleston, M.D., F.R.S. 1877. 
Medium 8vo. 25J. 

Clarendon^s History of the Rebellion and Civil Wars in 

England. 7 vols. 1839. i8mo. i/. \s. 

Clarendon^s History of the Rebellion and Civil Wars in 

England. Also his Life, written by 'himself, in which is included a Con- 
tinuation of his History of the Grand Rebellion. With copious Indexes. 
In one volume, royal Svo. 1842. i/. is, 

Clintons Epitome of the Fasti Hellenici. 1 85 1 . 8vo . 6s. 6d. 

Epitome of the Fasti Romani, 1854. 8vo. "js. 

Corpvs Poeticvm Boreale. The Poetry of the Old Northern 

Tongue, from the Earliest Times to the Thirteenth Century. Edited, clas- 
sified, and translated, with Introduction, Excursus, and Notes, by Gudbrand 
Vigfiisson, M.A., and F. York Powell, M.A. a vols. 1883. 8vo. 42J. 



CLARENDON PRESS, OXFORD. 1 1 



Freeman {E, A.). History of the Norman Conquest of Eng- 
land; its Causes and Results. In Six Volumes. 8vo. 5/. 9^. 6</. 

The Reign of William Rufus and the Accession of 



Henry the First. 2 vols. 8vo. i/. i6j. 

Gascoign^s Theological Dictionary ("Liber Veritatum ") ; 

Selected Passages, illustrating the condition of Church and State, I405-I458. 
With an Introduction by James £. Thorold Rogers, M.A. Small 4to. 
loj. 6d. 

Johnson {Samuel^ LL.D,), Boswetts Life of; including 

Bosweirs Journal of a Tour to the Hebrides, and Johnson's Diary of a 
Journey into North Wales. Edited by G. Birkbeck Hill, D.C.L. In six 
volumes, medium 8vo. With Portraits and Facsimiles of Handwriting. 
Half bound, 3/. 3^. (See p. 21.) 

Magna Carta, a careful Reprint. Edited by W. Stubbs, D.D. 

X879. 4to. stitched, IX. 

Passio et Miracula Beati Olaui. Edited from a Twelfth- 
Century MS. in the Library of Corpus Christi College, Oxford, with an 
Introduction and Notes, by Frederick Metcalfe, M.A. Small 4to. stiff 
covers, 6j. 

Protests of the Lords, including those which have been ex- 
punged, from 1624 to 1874; with Historical Introductions. Edited by James 
£. Thorold Rogers, M.A. 1875. 3vols. 8vo. 2/. 2x. 

Rogers {% E. 71). History of Agriculture and Prices in 

England, a.d. i 259-1 793. 

Vols. I and II (i 259-1400). 1866. 8vo. a/. 2/. 

Vols. Ill and IV (1401-1582). 1882. 8vo. 2/. lox. 

Vols. V and VI (1583- 1702). 8vo. 2/. loj. Just Published. 

The First Nine Years of the Bank of England. 8vo. 8j. 6d, 

Saxofi Chronicles {Two of the) parallel, with Supplementary- 
Extracts from the Others. Edited, with Introduction, Notes, and a Glos- 
sarial Index, by J.Earle, M.A. 1865. 8vo. i6x. 

Stubbs ( W.y D.D,\ Seventeen Lectures on the Study of 

Medieval and Modem History, &c., delivered at Oxford 1867-1884. Crown 
8vo. 8j. 6^. 

Sturlunga Saga, including the Islendinga Saga of Lawman 

Sturla Thordsson and other works. Edited by Dr. Gudbrand Vigflisson. 
In 2 vols. 1878. 8vo. 2/. 2j. 

York Plays, The Plays performed by the Crafts or Mysteries 

of York on the day of Corpus Christi in the 14th, 15th, and i6th centuries* 
Now first printed from the unique MS. in the Library of Lord Ashburnham. 
Edited with Introduction and Glossary by LucyToulmin Smith. 8vo. 2ix. 






12 CLARENDON PRESS, OXFORD. 

-^ — ♦ 

Manuscript Materials relating to the History of Oxford, 

Arranged by F. Madan, M.A. 8vo. *js. 6d. 

Statutes made for the University of Oxford^ and for the Colleges 

and Halls therein, by the University of Oxford Commissioners. 188 a. 8yo. 
I ax. 6^. 

Statuta Universitatis Oxoniensis. 1887. 8vo. 5j. 

The Oxford University Calendar for the year 1888. Crown 

8vo. 4/. (id. 

The present Edition includes all Class Lists and other University distinctions 

for the eight years ending with 1887. 

Also, sapplementary to the above, price Ss. (pp. 606), 

The Honours Register of the University of Oxford, A complete 

Record of University Hononrs, Officers, Distinctions, and Class Lists; of the 
Heads of Colleges, &c., &c., from the Thirteenth Century to 1883. 

The Examination Statutes for the Degrees of B,A,i B. Mus,, 

B.CL.^ and B.M. Revised to the end of Michaelmas Term, 1887. 8yo. 
sewed, u. 

The Students Handbook to the University and Colleges of 

Oxford, Ninth Edition. Crown 8vo. aj. 6^. 



MATHEMATICS, FHYSICAIi SCIENCE, A;o. 

Acland{H. W., M.D., F.R.S.). Synopsis of the Pathological 

Series in the Oxford Museum, 1867. 8vo. 2s, 6d, 

Bur don-Sander son (y., M.D,, F.R,SS. L, and E,), Transla- 
tions of Foreign Biological Memoirs, L Memoirs on the Physiology of Nerve, 
of Muscle, and of the Electrical Organ. Medium 8vo. au. 

De Bary {Dr. A.), Comparative Anatomy of the Vegetative 

Organs of the Phanerogams and Ferns, Translated and Annotated by F. O. 
Bower, M.A., F.L.S., and D. H. Scott, M.A., Ph.D., F.L.S. With 2^\ 
woodcuts and an Index. Royal 8vo., half morocco, i/. ax. 6</. 

Goebel (Dr. AT.). Outlines of Classification and Special Mor- 
phology of Plants. A New Edition of Sachs' Text-Book of Botany, Book IL 
English Translation by H. E. F. Gamsey, M.A. Revised by I. Bayley Balfoor, 
M.A., M.D., F.R.S. With 407 Woodcuts. Royal 8vo. half morocco, au. 

Sachs [Julius von). Lectures on the Physiology of Plants, 

Translated by H. Marshall Ward, M.A. With 445 Woodcuts. Royal 8vo. 
half morocco, i/. lis, 6d. 



CLARENDON PRESS, OXFORD, 13 



De Bary (Dr. A.). Comparative Morptiology and Biology of 

the Fungi, Myceiozoa and Bacteria. Authorised English Translation by 
Henry E. F. Ganisey, M.A. Revised by Isaac Bayley Balfour, M.A., M.D., 
F.R.S. With 198 Woodcuts. Royal 8vo., half morocco, i/. 2x. W. 

Lectures on Bacteria. Second improved edition. Au- 
thorised translation by H. K F. Gamsey, M.A. Revised by Isaac Bayley 
Balfour, M.A., M.D., F.R.S. With 30 Woodcuts. Crown 8vo. 6j. 



Annals of Botany. Edited by Isaac Bayley Balfour, M.A., 

M.D., F.R.S., Sydney H. Vines, D.Sc, F.R.S., and William Gilson Farlow, 
M.D., Professor of Cryptogamic Botany in Harvard University, Cambridge, 
Mass., U.S. A., and other Botanists. Royal 8vo. Vol. I. Now ready. 

Vol. 1. No. I, 8j. &/. No. a, *is. 6d, Nos. 3 and 4, i8s. 

Mailer (y.). On certain Variations in the Vocal Organs of 

the Passeres that have hitherto escaped notice. Translated by F. J. Bell, B.A .. 
and edited, with an Appendix, by A. H. Garrod, M. A., F.R.S. With Plates 
1878. 4to. paper covers, 7/. ^d. 

Price {Bartholomew^ M.A.y F.R.S.). Treatise on Infinitesimal 

Calculus. 

Vol.1. Differential Calculus. Second Edition. 8vo. 14/. 6^. 

Vol. II. Integral Calculus, Calculus of Variations, and Differential Equations. 
Second Edition, 1865. 8vo. i8x. 

Vol. III. Statics, including Attractions ; Dynamics of a Material Particle. 
Second Edition, 1868. 8vo. i6x. 

Vol. IV. Dynamics of Material Systems; together with a chapter on Theo* 
retical Dynamics, by W. F. Donkin, M.A., F.R.S. i86a. 8vo. i6j. 

Pritchard (C., D.D.^ F.R.S.). Uranometria Nova Oxoniensis. 

A Photometric determination of the magnitudes of all Stars visible to the naked 
eye, from the Pole to ten degrees south of the Equator. 1885. Royal 8vo. 
8j.6</. 

Astronomical Observations made at the University 

Observatory, Oxford, under the direction of C. Pritchard, D.D. No. i. 
1878. Royal 8vo. paper covers. 3/. 6d. 

Rigaud's Correspondence of Scientific Men of the i*jth Century ^ 

with Table of Contents by A. de Morgan, and Index by the Rev. J. Rigand, 
M.A. a vols. 1841-1802. 8vo. i&r. 6^. 

Rolleston {George, MJ>.y F.R.S. ). Forms of Animal Life. 

A Manual of Comparative Anatomy, with descriptions of selected types. 
Second Edition. Revised and enlarged by W. Hatchett Jackson, M.A. 
Medium, 8vo. doth extra, i/. 16s. 

Scientific Papers and Addresses. Arranged and Edited 

by William Turner, M.B., F.R.S. With a Biographical Sketch by Edward 
Tylor, F.R.S. With Portrait, Plates, and Woodcuts, a vols. 8vo. J/. 4X. 



1 4 CLARENDON PRESS, OXFORD. 



Westwood {J. O,, M.A„ F.RS.). TJiesaurus Entomologicus 

ffopeianusj or a Description of the rarest Insects in the Collection given to 
the University by the Rev. William Hope. With 40 Plates. 1874. Small 
folio, half morocco, 7/. lOJ. 



®&e Sbamtr 3Soofis of tj^ Cast. 

Translated by various Oriental Scholars, and^edttbd by 

F. Max MtJLLER. 

[Demy 8vo. cloth.] 

Vol. I. The Upanishads. Translated by F. Max Miiller. 

Part I. The Avi&ndogya-upanishad, The Talavak&ra-upanishad, The Aitareya- 
ftraffyak.i, The Kaushitaki-brihrna^a-upanishad, and The Vd.,^asaneyi-samhitd- 
upanishad. loj. 6d, 

Vol. II. The Sacred Laws of the Aryas, as taught in the 

Schools of Apastamba, Gautama, V&sish/^, and Baudh&yana. Translated by 
Prof. Georg Biihler. Part I. Apastamba and Gantama. loj. 6d. 

Vol. III. The Sacred Books of China. The Texts of Con- 
fucianism. Translated by James Legge. Part I. The Shfl King, The Reli- 
gions portions of the Shih King, and The Hsiio King. 1 2s. 6d, 

Vol. IV. The Zend-Avesta. Translated by James Darme- 

steter. Parti. TheVendtdid. iox.6^. 

Vol. V. The Pahlavi Texts. Translated by E. W. West. 

P^rt I. The Bondahlr, Bahman Ya^t, and Shayast 14-sh£yast. 12s. 6d, 

Vols. VI and IX. The Qur'dn. Parts I and 11. Translated 

by £. H. Palmer. 21s, 

Vol. VII. The Institutes of Vish«u. Translated by Julius 

Jolly. Tox. 6d, , 

Vol. VIII. The Bhagavadgita, with The Sanatsu^tiya, and 

The Anagtt&. Translated by K&shinfith Trimbak Telang. loj. 6^. 

Vol. X. The Dhammapada, translated from Pili by F. Max 

Miiller; and The Sutta-NipEta, translated from P&li by V. FausboU; being 
Canonical Books of the Bnddhists. lor. 6d, 



. CLARENDON PRESS, OXFORD, 15 

Vol. XI. Buddhist Suttas. Translated from Pdli by T. W, 

Rhys Davids, i. The Mah^parinlbb&na Snttanta ; a. The Dhamma-i&akka- 
ppavattana Sntta ; 3. The Tevi^^ Sattanta ; 4. The Akankheyya Sutta ; 
5. The/TetokhilaSatta; 6. The Mahd-sudassana Suttanta ; 7. The Sabbfisava 
Sutta. 10s. 6d. 

Vol. XII. The ^atapatha-Brihma^a, according to the Text 

of the M&dhyandina School. Translated by Julias Eggeling. Part I. 
Books I and II. 1 2x. 6^. 

Vol. XIII. Vinaya Texts. Translated from the P41i by 

T. W. Rhys Davids and Hermann Oldenbeig. Part I. The Pdtimokkha. 
The Mahftvagga, I-IV. loj. 6d. 

Vol. XIV. The Sacred Laws of the Aryas, as taught in the 

Schools of Apastamba, Gautama, VdsishMa and Baudh&yana. Translated 
by Georg Buhler« Part II. Vftsish/Aa and Baudh&yana. los. 6d, 

Vol. XV. The Upanishads. Translated by F. Max Muller. 

Part II. The Ka/>fa-upanishad, The Mu»^iSEika-npanishad, The Taittirtyaka- 
npanishad, The Bfihadftia^yaka-npanishad, The 5'vetaxvatara-upanishad, The 
Prar^a-npanishad, £nd The Maitr&ya^fa-BrdhmaMa-upanishad. loj. 6d. 

Vol. XVI. The Sacred Books of China. The Texts of 

Confucianism. Translated by James Legge. Part II. The Yt King. 
I or. 6d, 

Vol. XVII. Vinaya Texts. Translated from the P4H by 

T. W. Rhys Davids and Hermann Oldenberg. Part II. The Mahdvagga, 
V-X. The A'ulUvagga, I-III. ios,6d. 

Vol. XVIII. Pahlavi Texts. Translated by E. W. West. 

Part II. The Dd/Astftn-t Dtntk and The Epistles of M&ndryHhar. 1 2s. 6d, 

Vol. XIX. The Fo-sho-hing-tsan-king. A Life of Buddha 

by Axvaghosha Bodhisattva, translated from Sanskrit into Chinese by 
Dharmaraksha, aj>. 420, and from Chinese into English by Samuel Beal. 
xor. 6d, 

Vol. XX. Vinaya Texts. Translated from the P41i by T. W. 

Rhys Davids and Hermann Oldenberg. Part III. The Aldlavagga^IV-XII. 
loj. 6d, 

Vol. XXI. The Saddharma-pu;?^arika ; or, the Lotus of the 

True Law. Translated by H. Kem. laj. 6d. 

Vol. XXII. G^aina-Satras. Translated from Prakrit by Her- 
mann Jacobi. Parti. The AHrdnga-Siltra. The Kalpa-Sdtnu lox. 6</. 



i6 CLARENDON PRESS, OXFORD. 

Vol. XXIII. The Zend-Avesta. Translated by James Dar- 

mesteter. Part II. The Strdzahs, Yajts, and Nyftyir. los. 6d, 

Vol. XXIV. Pahlavi Texts. Translated by E. W. West, 

Part III. Dtn&-t Matndg-t Khirad, 51kand.g(imfintk, and Sad-Dar. 
losx. 6d. 



Seoond Series. 

Vol. XXV. Manu. Translated by Geor^ Biihler. 21s, 
Vol. XXVI. The 5atapatha-Br4hma«a. Translated by 

Julius Eggeling. Part II. 12s. 6d. 

Vols. XXVII and XXVIII. The Sacred Books of China. 

The Texts of Confucianism. Translated by James Legge. Parts III and IV, 
The Lt Al| or Collection of Treatises on the Rules of Propriety, or Ceremonial 
Usages. 2$s. 

Vols. XXIX and XXX. The GrAya-Sfttras, Rules of Vedic 

Domestic Ceremonies. Translated by Hermann Oldenberg. 

Part I (Vol. XXIX), 12s, 6d. fust Published, 
Part II (Vol. XXX). In the Press, 

Vol. XXXI. The Zend-Avesta. Part III. The Yasna, 

Visparad, Afrtnag&n, and Gihs. Translated by L. H. MUls. 12s. 6d. 



The following Volumes are in the Press :— 

Vol. XXXII. Vedic Hymns. Translated by F. Max Mullen 

Parti. 

Vol. XXXIII. N4rada, and some Minor Law-books. 

Translated by Julius Jolly. IPreparing'.'] 

Vol. XXXIV. The Veddnta-Sfttras, with 5ankara's Com- 
mentary. Translated by G. Thibaut [Preparing-.'] 

♦«♦ The Second Series will consist 0/ jTwenty-Foisr Volumes, 



CLARENDON PRESS, OXFORD, 17 



€hxtxi^m |pw88 %mi%. 



I. ENQLISH, A;o. 
An Elementary English Grammar and Exercise Book. By 

O. W. Tancock, M.A. Second Edition. Extra fcap. 8vo. is. 6d. 

An English Grammar and Reading Book^ for Lower Forms 

in Classical Schools. By O. W. Tancock, M.A. Foorth Edition. Extra 
fcap. 8vo. 5x. 6d. 

Typical Selections from the best English Writers^ with Intro- 
ductory Notices. Second Edition. In 2 vols. Extra fcap. 8vo. 3/. 6^. each. 
Vol. I. Latimer to Berkeley. Vol. II. Pope to Macaolay. 

Shairp {J. C, LL.D.). Aspects of Poetry; being Lectures 

delivered at Oxford. Crown 8vo. lOf. dd. 



A Book for the Beginner in Anglo-Saxon, By John Earle, 

M.A. Third Edition. Extra fcap. 8vo. 2x. 6d. 

An Anglo-Saxon Reader. In Prose and Verse. With Gram- 
matical Introduction. Notes, and Glossary. By Henry Sweet, M.A. Fourth 
Edition, Revised and Enlarged. Extra fcap. 8vo. &r. 6d. 

A Second Anglo-Saxon Reader. By the same Author. Extra 

fcap. 8vo. 4r. 6^. 

An Anglo-Saxon Primer ^ with Grammar, Notes y and Glossary. 

By the same Author. Second Edition. Extra fcap. 8yo. 2j. 6d, 

Old English Reading Primers; edited by Henry Sweet, M.A. 

I. Selected Homilies of iGlfric. Extra fcap. Svo., stiff covera, is. 6d. 

II. Extracts from Alfred*s Orosius. Extra fcap. 8vo., stiff covers, is, 6d, 

First Middle English Primer^ with Grammar and Glossary. 

By the same Author. Extra fcap. 8vo. ax. 

Second Middle English Primer. Extracts from Chaucer, 

with Granmiar and Glossary. By the same Author. Extra fcap. 8vo. 2J. 

A Concise Dictionary of Middle English, from A.D. 11 50 to 

1580. By A. L. Mayhew, M.A., and W. W, Skeat, LittD. Crown 8vo. 
half roan, 7^. 6^. 

A Handbook of Phonetics, including a Popular Exposition of 

the Principles of Spelling Reform. By H. Sweet » M.A. Ext. fcap. 8vo. 41. 6</. 

C 



18 CLARENDON PRESS, OXFORD. 

Elementarbtich des Gesprochenen Englisch. Grammatiky 

Textc nnd Glossar. Von Henry Sweet. Second Edition, Extra fcap. 8vo., 
itiff covers, ax. 6</. 

History of English Sounds from the earliest Period. With 

full Word-Lists. By Henry Sweet, M,A. Demy 8vo. 14?. 

Principles of English Etymology. First Series. The Native 

Element. By W. W. Skeat, Litt.D. Crown 8vo. 9X. 

The Philology of the English Tongue. By J. Earle, M.A. 

Fourth Edition. Extra fcap. 8vo. 7/. 6</. 

An Icelandic Primer^ with Grammar, Notes, and Glossary. 

By Henry Sweet, M.A. Extra fcap. 8vo. 3J. W. 

An Icelandic Prose Reader, with Notes, Grammar, and Glossary. 

By G. Vigfiisson, M.A., and F. York Powell, M.A. Ext. fcap. 8vo. ioj. dJ. 

The Ormulum; with the Notes and Glossary of Dr. R. M. 

White. Edited by R. Holt, M.A. 1878. a vols. Extra fcap. 8vo. a ix. 

Specimens of Early English. A New and Revised Edition. 

With Introduction, Notes, and Glossarial Index. 

Part I. By R. Morris, LL.D. From Old English Homilies to King Horn 
(A. D. 1 1 50 to A.D. 1500). Second Edition. Extra fcap. 8vo. 9^. 

Part II. By R. Morris, LL.D., and W. W. Skeat, Litt.D. From Robert 
of Gloucester to Gower (A.D. 1398 to A.D. 1393). Third Edition. 
Extra fcap. 8vo. *is. td. 

Specimens of English Literature, from the 'Ploughmans 

Crede' to the • Shepheardes Calender' (a.d. 1394 to a.d. 1579). With Intro- 
duction, Notes, and Glossarial Index. By W. W. Skeat, Litt.D. Fourth 
Edition. Extra fcap- 8vo. yx. 6^. 



The Vision of William concerning Piers the Plowman, in three 

Parallel Texts ; together with Richard the Redeless. By William Langland 
(about 1363-1399 A.D.). Edited from numerous Manuscripts, with Preface, 
Notes, and a Glossary, by W. W. Skeat, Litt.D. 3 vols. 8yo. 3Xj. 6</. 

The Vision of William concerning Piers t/ie Plowman, by 

William Langland. Edited, with Notes, by W. W. Skeat, Litt.D. Fourth 
Edition Extra fcap. 8yo- 41. 6^. 

Chaucer. I. The Prologue to the Canterbury Tales; the 

Knightes Tale; The Nonne Prestes Tale. Edited by R. Morris, LL.D. 
Sixty-sixth thousand. Extra fcap. 8vo, 2s. 6d. 

II. The Prioresses Tale; Sir Thopas ; T/ie Monkes 

Tale ; The Clerkes Tale ; The Squieres Tale, &c. Edited by W. W. Skeat, 
LittD. Third Edition. Extra leap. 8yo. ^. 6d, 



CLARENDON PRESS, OXFORD. 19 

Chaucer. III. The Tale of the Man ofLawe ; The Pardoneres 

Tale ; The Second Nonnes Tale ; The Chanotms Yemaimes Tale. By the 
game Editor. New Edition, Rtvised. Extra fcap. 8vo. 4/. 6</. 

Gamelyn, The Tale of. Edited with Notes, Glossary, &c., by 

W. W. Skeat, Litt.D. Extra fcap. 8vo. Stiff covers, \s. 6d. 

Minot {Laurence). Poems. Edited, with Introduction and 

Notes, by Joseph Hall, M.A., Head Master of the Hnlme Grammar School, 
Manchester. Extra fcap. 8vo. 41. 6d. 

Spenser^ s Faery Queene. Books I and II. Designed chiefly 

for the use of ISchools. With Introduction and Notes by G. W^ Kitchin, D.D., 
and Glossary by A. L. Mayhew, M.A. Extra fcap. 8vo. 2s. fid. each. 

Hooker. Ecclesiastical' Polity, Book I. Edited by R, W. 

Chnrch, M.A. Second Edition. Extra fcap. 8vo. ax. 

OLD ENGLISH DRAMA. 
The Pilgrimage to Parnassus with The Two Parts of the 

Return from Parnassus, Three Comedies performed in St. John's College, 
Cambridge, A.D. mdxcvii-mdci. Edited from MSS. by the Rev. W. 1>. 
Macray, M.A., F.S.A. Medinm 8vo. Bevelled Boards, Gilt top, Sx. 6d. 

Marlowe and Greene. Marlowe*s Tragical History- of Dr. 

Faustus^ and Greene's Honourable History of Friar Bacon and Friar Bungay. 
Edited by A. W. Ward, M.A. Hew and Enlarged Edition. Extra fcap. 
8vo. 6j. 6d. 

Marlowe. Edward ll. With Introduction, Notes, &c. By 

O. W. Tancock, M.A. Extra fcap. 8vo. Paper covers, a/. Cloth 3X. 

SHAKESPEARE. 
Shakespeare. Select Plays. Edited by W. G. Clark, M.A., 

and W. Aldis Wright, M.A. Extra fcap. 8vo. stiff covers. 
The Merchant of Venice, is. Macbeth, is. 6d, 
Richard the Second, is. 6d. Hamlet. 2s. 

Edited by W. Aldis Wright, M.A. 

The Tempest, is, 6d. Midsummer Night's Dream, is. 6d. 

As You Like It. is. 6d. Coriolanus. aj. 6d. 

Julius Caesar, a/. Henry the Fifth, is. 

Richard the Third, aj. 6d. Twelfth Night, is. 6d. 

King Lear. is. 6d. King John. is. 6d. 

Shakespeare as a Dramatic Artist; a popular Illustration of 

the Principles of Scientific Criticism. By R. G. Moulton, M.A. Crown 8vo. 5/. 

c a 



ao CLARENDON PRESS, OXFORD. 

Bacon, I. Advancement of Learning. Edited by W, Aldis 

Wright, M.A. Third Edition. Extra fcap. Svo. 41. 6d. 

II. The Essays. With Introduction and Notes. By 



S. H. Reynolds, M.A., late Fellow of Brasenose College. In Preparation. 

Milton. I. Areopagitica. With Introduction and Notes. By 

John W. Hales, M.A. Third Edition. Extra fcap. 8vo. 3J. 

II. Poems. Edited by R. C. Browne, M.A. % vols. 

Fifth Edition. Extra fcap. 8vo. 6j. 6d, Sold separately. Vol. 1. 4J. ; Vol. II. 3X. 

In paper oovera : — 
Lycidas, id. L' Allegro, id. II Penseroso, 4^. Gomus, Sd. 

III. Paradise Lost. Book I. Edited by H. C. Beeching. 



Extra fcap. 8vo. stiff cover, u. 6^. ; in white Parchment, 3/. 6d. 

IV. Samson Agonistes. Edited with Introduction and 

Notes by John Churton Collins. Extra fcap. Svo. stiff covers, \s, 

Bunyan. I. The Pilgrim's Progress, Grace Abounding, Rela-- 

iion of the Imprisonment of Mr, John Bunyan, Edited, with Biographical 
Introduction and Notes, by E. Venables, M.A. 1879. Extra fcap. Svo. 5x. 
In ornamental Parchment, 6x. 

II. Holy War, 6f*c. Edited by E. Venables, M.A. 



In the Press. 

Clarendon. History of the Rebellion. Book VI. Edited 

by T. Arnold, M.A. Extra fcap. Svo. 4^. 6d. 

Dryden. Select Poems. Stanzas on the Death of Oliver 

Cromwell ; Astraea Redux ; Annus Mirabilis ; Absalom and Achitophel ; 
Religio Laici ; The Hind and the Panther. Edited by W. D. Christie, M.A. 
Second Edition. Extra fcap. 8vo. 31. 6^. 

Locke's Conduct of the Understanding. Edited, with Intro- 
duction, Notes, &c., by T. Fowler, D.D. Second Edition. Extra fcap. Svo. ax. 

Addison. Selections from Papers in the Spectator. With Notes. 

By T. Arnold, M.A. Extra fcap. Svo. 4^. 6J. In ornamental Parchment, 6j. 

Steele. Selections from the Tatler, Spectator, and Guardian. 

Edited by Austin Dobson. Extra fcap. Svo. \s. 6d. In white Parchment, *js, 6d. 

Pope. With Introduction and Notes. By Mark Pattison, B.D. 

I. Essay on Man.^ Extra fcap. Svo. u. 6d. 

II. Satires and Epistles. Extra fcap. Svo. %s. 



>f 



CLARENDON PRESS, OXFORD. ai 

PamelL The Hermit. Paper covers, %d. 

Gray. Selected Poems. Edited by Edmund Gosse. Extra 

fcap. 8vo. Stiff covers, \s. 6d, In white Parchment, 3/. 

EUgy and Ode on Eton College. Paper covers, %d. 

Goldsmith. Selected Poems. Edited, with Introduction and 

Notes, by Austin Dobson. Extra fcap. 8yo. 3J. 6d, In white Parchment, 

The Deserted Villc^e. Paper covers, %d. 

Johnson. I. Rcuselas; Lives of Dry den and Pope. Edited 

bv Alfred Milnes, M.A. (London). Extra fcap. 8yo. 41. td,, or Livis of 
Drydifi and Pope only, stiff covers, a/. 6</. 

II. Rasselas. Edited, with Introduction and Notes, by 

G.BirkbeckHiIl,D.C.L. Extra fcap. 8vo. Bevelled boards, 3/. 6</. In white 
Parchment, \s, M, 

III. Vanity of Human Wishes. With Notes, by E. J. 

Payne, M.A. Paper covers, \i. 

IV. Life of Milton. By C. H. Firth, M.A. In the Press. 



V. Wit and Wisdom of Samuel Johnson. Edited by 

G. Birkbeck Hill, D.C.L. Crown 8vo. 71. 6d, 

VI. BosweUs Life of Johnson. With the Journal of a 

Tour to the Hebrides, Edited, with coplons Notes, Appendices, and Index, by 
G. Birkbeck Hill, D.C.L., Pembroke College. With PortraiU and Facsimiles. 
6 vols. Medimn 8vo. Half bound , 3/. 3^. 

Cowper. Edited, with Life, Introductions, and Notes, by 

H. T. Griffith, BA. 

I. The Didactic Poems of lyS^^ with Selections from the 

MinorPieces,A.D. 1779-1783. Extra fcap. 8vo. 3/. 

II. The Task, with Tirocinium^ and Selections from the 



Minor Poems, A.D. 1 784-1 799. Second Edition. Extra fcap. 8vo. 3/. 

Burke. Select Works. Edited, with Introduction and Notes, 

by E. J. Payne, M.A. 

- I. Thoughts on the Present Discontents ; the two Speeches 

OH America* Second Edition. Extra fcap. 8vo. 41. td. 

11. Reflections on the French Revolution. Second Edition. 

Extra fcap. 8vo. 5/. 

III. Four Letters on the Proposals for Peace with the 

Regidde Directory of France. Second Edition. Extra fcap. 8vo. 5^. 



32 CLARENDON PRESS, OXFORD. 



Keats. Hyperion, Book I. With Notes by W.T. Arnold, B- A. 

Paper covers, 4^. 

Byron. Childe Harold. Edited, with Introduction and Notes, 

by H. F. Tozer, M.A. Extra fcap. 8vo. jj. (id. In white Parchment, 5X. 

Scott. Lay of the Last Minstrel. Edited with Preface and 

Notes by W. Minto, M.A. With Map. Extra fcap. 8vo. Stiff covers, ax. 
Ornamental Parchment, ^. 6^. 

Lay of the Last Minstrel. Introduction and Canto I, 

with Preface and Notes, by the same Editor. 6d. 

II. LATIN. 
Rudimenta Latina. Comprising Accidence, and Exercises of 

a very Elementary Character, for the use of Beginners. By John Barrow 
Allen, M.A. Extra fcap. 8vo. 3J. 

An Elementary Latin Grammar. By the same Author. 

Fifty-Seventh Thousand. Extra fcap. 8vo. is.dd, 

A First Latin Exercise Book. By the same Author. Fourth 

Edition. Extra fcap. 8vo. 2j. dd. 

A Second Latin Exercise Book. By the same Author. Extra 

fcap. 8vo. 3J. ^d. 

A Key to First and Second Latin Exercise Books, in one volume, price 5^. 
Supplied to Teachers only on application to the Secretary of the Cuirendon 
Press. 

Reddenda Minora, or Easy Passages, Latin and Greek, for 

Unseen Translation. For the use of Lower Forms. Composed and selected 
by C. S. Jerram, M.A. Extra fcap. 8vo. \s. 6d, 

Anglice Reddenda, or Extracts, Latin and Greek, for 

Unseen Translation. By C. S. Jerram, M.A. Third Edition, Revised and 
Enlarged. Extra fcap. 8vo. 2j. 6^. 

Anglice Reddenda. Second Series. By the same Author. 

Extra fcap. 8vo. 3^. 

Passages for Translation into Latin. For the use of Passmen 

and others. Selected by J. Y. Sargent, M.A. Seventh Edition. Extra fcap. 
8vo. 2 J. 6</. 

Exercises in Latin Prose Composition; with Introduction, 

Notes, and Passages of Graduated Difficulty for Translation into Latin. By 
G. G. Ramsay, M.A., LL.D. Second Edition. Extra fcap. 8vo. 4^. 6</. 

Hints and Helps for Latin Elegiacs. By H. Lee- Warner, M.A. 

Extra fcap. 8vo. 3J. dd. 

First Latin Reader. By T. J. Nunns, M.A. Third Edition. 

Extra fcap. 8vo. 2s. 



CLARENDON PRESS, OXFORD. 23 

Caesar. The Commentaries (for Schools). With Notes and 

Maps. By Charles £. Moberly, M.A. 

Part I. Thi Gallic War. Second Edition. Extra fcap. 8vo. 4/. 6d, 
Part n. The Civil War, Extra fcap. 8vo. 3s. 6rf. 

Thi Civil War. Book I. Second Edition. Extra fcap. 8to. 2s. 

Cicero. Speeches against Catilina. By E. A. Upcott, M.A., 

Assistant Master in Wellington College. In one or two Parts. Extra fcap. 
8vo. 2x. 6</. 

Cicero. Selection of interesting and descriptive passages. With 

Notes. By Henry Walford, M.A. In three Parts. Extra fcap. 8 vo. 41. 6</. 

Each Part separately, limp, is, 6d. 

Part I. Anecdotes from Grecian and Roman History. Third Edition 
Part II. Omens and Dreams: Beauties of Nature. Third Edition. 
Part III. Rome*s Rule of her Provinces. Third Edition. 

Cicero. De Senectute. Edited, with Introduction and Notes, 

by L. Huxley, M.A. In one or two Parts. Extra fcap. 8yo. 3/. 

Cicero. Selected Letters (for Schools). With Notes. By the 

late C. £. Prichard, M.A., and E. R. Bernard, M.A. Second Edition. 
Extra fcap. 8vo. 31. 

Cicero. Select Orations (for Schools). In Verrem I. De 

Imperio Gn. Pompeii. Pro Archia. Philippica IX. With Introduction and 
Notes by J. R. King, M.A. Second Edition. Extra fcap. 8vo. 3j. 6^. 

Cicero, In Q, Caecilium Divinatio, and In C. Verrem Actio 

Prima, With Introduction and Notes, by J. R. King, M.A. Extra fcap. 8vo. 
limp, i^. td, 

Cicero. Speeches against Catilina. With Introduction and 

Notes, by E. A. Upcott, M.A. In one or two Parts. Extra fcap. 8vo. 
2X. td. 

Cornelius Nepos. With English Notes. By Oscar Browning, 

M.A. Third Edition. Revised by W. R. Inge, M.A. (In one or two Parts.) 
Extra fcap. 8vo..3j. 

Horace. Selected Odes. With Notes for the use of a Fifth 

Form. By E. C. Wickham, M.A. In one or two Parts. Extra fcap. Svo. 
cloth, 2S. 

Livy. Selections (for Schools). With Notes and Maps. By 

H. Lee- Warner, M.A. Extra fcap. Svo. In 'Parts, limp, each \s, 6d, 

Part I. The Caudine Disaster. Part II. Hannibars Campaign 
in Italy. Part III. The Macedonian War. 

Livy. Books V-VII. With Introduction and Notes. By 

A. R. Cluer, B.A. Second Edition. Revised by P. £. Matheson, M.A. 
(In one or two Parts.) Extra fcap. Svo. 5J. 

Livy. Books XXI, XXII, and XXIII. With Introduction 

and Notes. By M. T. Tatham, M.A. Extra fcap. Svo. 4;. 6d, 



I 



a4 CLARENDON PRESS, OXFORD, 

Ovid. Selections for the use of Schools. With Introductions 

and Notes, and an Appendix on the Roman Calendar. By W. Ramsay, Inf. A. 
Edited by G. G. Ramsay, M.A. Thiid Edition. Extra fcap. 8to. 5/. <S^. 

Ovid* Tristia. Book I. The Text revised, with an Intro- 
duction and Notes. By S. G. Owen, B.A. Extra fcap. 8vo. 3J. 6^. 

Plautus. Captivu Edited by W. M. Lindsay, M.A. Extra 

fcap. 8vo. (In one or two Parts.) 2J. 6</. 

Plautus. The Trinummus. With Notes and Introductions. 

(Intended for the Higher Forms of Public Schools.) By C. E. Freeman, M. A., 
and A. Sloman, M.A. Extra fcap. 8vo. 31. 

Pliny. Selected Letters (for Schools). With Notes. By the 

late C. E. Prichard, MA., and E. R. Bernard, M.A. Extra fcap. 8yo. 31. 

Sallust. With Introduction and Notes. By W. W. Capes, 

M.A. Extra fcap. 8vo. 41. 6d, 

Tacitus. The Annals. Books I-IV. Edited, with Introduc- 
tion and Notes (for the use of Schools and Junior Students), by H. Fumeaux, 
M.A. Extra fcap. 8vo. 51. 

Tacittis. The Annals. Book I. With Introduction and Notes, 

by the same Editor. Extra fcap. 8vo. limp, 2s, 

Terence. Andria. With Notes and Introductions. By C. 

E. Freeman, M.A., and A. Sloman, M.A. Extra fcap. 8vo. 3^. 

Adelphi. With Notes and Introductions. (Intended for 

the Higher Forms of Public Schools.) By A. Sloman, M.A. Extra fcap. 
8vo. 3J. 

Phormio. With Notes and Introductions. By A. 



Sloman, M.A. Extra fcap. 8vo. 3/. 

Tibulltis and Proper tins* Selections. Edited by G. G. Ramsay, 

M.A. Extra fcap. 8yo. (In one or two vols.) dr. 

Virgil. With Introduction and Notes. By T. L. Papillon, 

M.A. Two vols. Crown 8vo. loj. 6^. The Text separately, 41. &/. 

Virgil. Bucolics. Edited by C. S. Jerram, M.A. In one 

or two Paits. Extra fcap. 8vo. 2s. 6^. 

Virgil. Aeneid I. With Introduction and Notes, by C. S. 

Jerram, M.A. Extra fcap. 8vo. limp, u. 6</. 

Virgil. Aeneid IX. Edited, with Introduction and Notes, 

by A. E. Haigh, M.A., late Fellow of Hertford College, Oxford. Extra 
fcap. 8vo. limp, u. (>d. In two Parts, 2j. 



CLARENDON PRESS, OXFORD. 25 

Avianus, The Fables of. Edited, with Prolegomena, Critical 

Appantus, Commentary, etc. By Robinson Ellis, M.A., LL.D. Demy 8vo. 

Catulli Veronensis Liber. Iterum recognovit, apparatum cri- 

ticnm prolegomena appendices addidit, Robinson Ellis, A.M. 1878. Demy 
8to. I 6s. 

A Commentary on Catullus. By Robinson Ellis, M.A. 

1876. DemySvo. idr. 

Catulli Veronensis Carmina Selecia, secmidum recognitionem 

Robinson Ellis, A.M. Extra fcap. Svo. 31. 6d. 

Cicero de Oratore. With Introduction and Notes. By A. S. 

Wilkins, M.A. 

Book I. Second Edition. 1888. 8vo. *js,6d. Book II. 1881. 8yo. 5/. 

Philippic Orations. With Notes. By J. R. King, M.A. 

Second Edition. 1879. 8yo. lor. 6d, 

Cicero. Select Letters. With English Introductions, Notes, 

and Appendices. By Albert Watson, M. A. Third Edition. Demy 8vo. i8s. 

Select Letters. Text. By the same Editor. Second 

Edition. Extra fcap. 8yo. 41. 

pro Cluentio. With Introduction and Notes. By W. 



Runsay, M.A. Edited by G. G. Ramsay, M.A. 2nd Ed. Ext. fcap. 8vo. 5/. 6d. 

Horace. With a Commentary. Volume I. The Odes, Carmen 

Secnlare, and Epodes. By Edward C. Wickham, M.A. Second Edition. 
1877. DemySvo. 12s. 

A reprint of the above, in a size suitable for the use 

of Schools. In one or two Parts. Extra fcap. 8vo. dr. 

Livy, Book I. With Introduction, Historical Examination, 

and Notes. By J. R. Seeley, M. A. Second Edition. 1881. 8vo. 6x. 

Ovid. P. Ovidii Nasonis Ibis* Ex Novis Codicibus edidit, 

Scholia Vetera Commentorinm cum Prolegomenis Appendice Indice addidit, 
R. EUis, A.M. 8vo. lor. 6d. 

Persius. The Satires. With a Translation and Commentary. 

By John Conington, M.A. Edited by Henry Nettleship, M.A. Second 
Edition. 1874. 8vo. *js, 6d. 

Juvenal. XIII Satires. Edited, with Introduction and 

Notes, by C. H. Pearson, M.A., and Herbert A. Strotig, M.A., LL.D., Professor 
of Latin in Liverpool University Coll^ie^ Victoria University. In two Parts. 
Crown 8vo. Complete, 6s. 

Also separately. Part I. Introdnction, Text, etc., y. Part II. Notes, 3/. 6^. 

Tacitus. The Annals. Books I- VI. Edited, with Intro- 
duction and Notes, by H. Fumeanx, M.A. 8vo. i&r. 



26 CLARENDON PRESS, OXFORD. 

King (y. E.y M.A.) and C. Cookson, M,A. The Principles of 

Sound and Inflexion^ as illustrated in the Greek and Latin Languages. i88S. 
8vo. 1 8 J. 

Nettleship (//,, MjlJ). Lectures and Essays on Subjects con- 
nected with Latin Scholarship and Literature. Crown Svo. 7j. 6</. 

The Roman Satura, Svo. sewed, is. 

Ancient Lives of Vergil. Svo. sewed, aj. 

Papillon {T. L., M.A.). A Manual of Comparative Philology. 

Third Edition, Revised and Corrected. 1882. Crown Svo. 6x. 

Pinder {Norths Af.A.). Selections from the less known Latin 

Poets. 1869. Svo. 15/. 

SeUar ( W. F., M,A.). Roman Poets of the Augustan A£^e. 

Virgil. New Edition. 1883, Crown Svo. 9J. 

Roman Poets of the Republic. New Edition, Revised 

and Enlarged. 1881. Svo. 141. 

Wordsworth (y., M.A,). Fragments and Specimens of Early 

Latin. With Introductions and Notes. 1S74. Svo. i8j. 

III. GBEEK. 

A Greek Primer^ for the use of beginners in that Language. 

By Charles Wordsworth, D.C.L. Seventh Edition. Extra fcap. Svo. is. 6d. 

A Greek Testament Primer. An Easy Grammar and Read- 
ing Book for the use of Students beginning Greek. By the Rev. E. Miller, 
M.A. Extra fcap. Svo. 3J. dd. 

Easy Greek Reader. By Evelyn Abbott, M.A. In one or 

two Parts. Extra fcap. Svo. 3J. 

Graecae Grammaticae Rudimenta in usum Scholarum. Auc- 

tore Carolo Wordsworth, D.C.L. Nineteenth Edition, 1883. i2mo. 41. 

A Greek-English Lexicon^ abridged from Liddell and Scott's 

4to. edition, chiefly for the use of Schools. Twenty-first Edition. 18S6. 
Square i2mo. 7^. dd, 

Greek Verbs, Irregular and Defective. By W. Veitch. Fourth 

Edition. Crown Svo. loj. 6</. 

The Elements of Greek Accentuation (for Schools) : abridged 

from his larger work by H. W. Chandler, M.A. Extra fcap. Svo. %s. 6d. 

A Series of Graduated Greek Readers: — 

First Greek Reader. By W. G. Rushbrooke, M.L. Second 

Edition. Extra fcap. Svo. ax. 6d, 

Second Greek Reader. By A. M. Bell, M.A. Extra fcap. 

Svo. 3J. 6d. 



CLARENDON PRESS, OXFORD. 27 

Fourth Greek Reader; being Specimens of Greek Dialects. 

With Introductions, etc. By W. W. Merry, D.D. Extra fcap. 8vo. 4/. 6d, 

Fifth Greek Reader. Selections from Greek Epic and 

Dramatic Poetiy, with Introductions and Notes. By Evelyn Abbott, M.A. 
Extra fcap. 8vo. 41. 6</. 

The Golden Treasury of Ancient Greek Poetry: being a Col- 
lection of the finest passages in the Greek Classic Poets, with Introductory 
Notices and Notes. By R. S. Wright, M.A. Extra fcap. 8vo. 8x. td. 

A Golden Treasury of Greek Prose, being a Collection of the 

finest passages in the principal Greek Prose Writers, with Introductory Notices 
and Notes. By R. S. Wright, M.A., and J. £. L. Shadwell. M.A.' Extrafcap. 
8vo. 4^. 6d. 

Aeschylus^ Prometheus Bound (for Schools), With Introduc- 
tion and Notes, by A. O. Prickard, M.A. Second Edition. Extra fcap. 8vo. aj. 

Agamemnon. With Introduction and Notes, by Arthur 

Sidgwick, M.A. Third Edition. In one or two parts. Extra fcap. 8vo. 3/. 

Choephoroi. With Introduction and Notes by the same 

^ditor. Extrafcap. 8vo. 31. 

Eumenides. With Introduction and Notes, by the same 

Editor. In one or two Parts. Extra fcap. Svo. 3/. 

Aristophanes, In Single Plays. Edited, with English Notes, 

Introductions, &c., by W. W. Merry, D.D. Extra fcap. 8vo. 
I. The Clouds, Second Edition, is, 
II. The Achamians, Third Edition.. In one or two parts, 31. 

III. The Frogs, Second Edition. In one or two parts, 3J. 

IV. The Knights. In one or two parts, y, 

Cebes. Tabula, With Introduction and Notes. By C. S. 

Jerram, M.A. Extra fcap. Svo. is, 6d, 

Demosthenes. Orations against Philip. With Introduction 

and Notes, by Evelyn Abbott, M.A., and P. E. Matheson, M.A. Vol. I. 
Philippic I. Olynthiacs I-III. In one or two Parts. Extra fcap. Svo. 3^. 

Euripides. Alcestis (for Schools), By C. S. Jerram, M.A, 

Extra fcap. 8vo. 2s,6d. 

Helena. Edited, with Introduction, Notes, etc., for 

Upper and Middle Forms. By C. S. Jerram, M.A. Extra fcap. Svo. 3/. 

Iphigenia in Tauris. Edited, with Introduction, Notes, 

etc., for Upper and Middle Forms. By C. S. Jerram, M.A. Extra fcap. Svo, 
cloth, 3/. 

Medea. By C. B. Heberden, M.A. In one or two Parts. 



Extra fcap. Svo. ax. 



88 CLARENDON PRESS, OXFORD. 

Herodotus, Book IX. Edited, with Notes, by Evelyn Abbott, 

M.A. In one or two Parts. Extra fcap. Svo. ^. 

Herodotus^ Selections from. Edited, with Introduction, Notes, 

and a Map, by W. W. Merry, D.D. Extra fcap. Svo. ts. 6d. 

Homer. Odyssey, Books I-XII (for Schools). By W. W. 

Merry, D.D. Fortieth Thousand. (In one or two Parts.) Extra fcap. 
Svo. 5J. 

Books I, and II, separately ^ each u, 6d. 

Odyssey, Books XIII-XXIV (for Schools). By the 

same Editor. Second Edition. Extra fcap. Svo. 5/. 

Hiad, Book I (for Schools). By D. B. Monro, M.A. 



Second Edition. Extra fcap. Svo. 2s. 

Iliad, Books I-XII (for Schools). With an Introduction, 

a brief Homeric Grammar, and Notes. By D. B. Monro, M.A. Second 
Edition. Extra fcap. Svo. 6j. 

Iliad, Books VI and XXI. With Introduction and 

Notes. By Herbert Hailstone, M.A. Extra fcap. Svo. is, 6d. each. 

Lucian. Vera Historia (for Schools). By C. S. Jerram, 

M.A. Second Edition. Extra fcap. Svo. u. 6tf. 

Lysias. Epitaphios. Edited, with Introduction and Notes, 

by F. J. Snell, B.A. (In one or two Parts.) Extra fcap. Svo. aj. 

Plato. Meno. With Introduction and Notes. By St. George 

Stock, M.A., Pembroke College. (In one or two Parts.) Extra fcap. Svo. 
2s. 6d. 

Plato. The Apology. With Introduction and Notes. By 

St. George Stock, M.A. (In one or two Parts.) Extra fcap. Svo. is. 6</. 

Sophocles. For the use of Schools. Edited with Intro- 
ductions and English Notes. By Lewis Campbell, M. A., and Evelyn Abbott, 
M.A. New and Revised Edition. 2 Vols. Extra fcap. Svo. lOJ. 6d. 

Sold separately, Vol. I, Text, 4s, 6d. ; Vol. II, Explanatory Notes, 6/. 

Sophocles. In Single Plays, with English Notes, &c. By 

Lewis Campbell, M.A., and Evelyn Abbott, M.A. Extra fcap. Svo. limp. 
Oedipus Tyrannus, Philoctetes. New and Revised Edition, 2/. each. 
Oedipus Coloneus, Antigone, is. ^d. each. 

Ajax, Electra, Trachiniae, 21. each. 

Oedipus Rex: DindorPs Text, with Notes by the 

present Bishop of St. David's. Extra fcap. Svo. limp, is, ^d. 

Theocritus (for Schools). With Notes. By H. Kynaston, 

D.D. (late Snow). Third Edition. Extra fcap. Svo. 4/. dd. 



CLARENDON PRESS, OXFORD. 89 



Xenophon. Easy Selections (for Junior Classes). With a 

Vocabulary, Notes, and Map. By J. S. Phillpotts, B.C.L,, and C. S. Jcrram, 
M. A. Third Edition. Extra fcap. Svo. 3/. id. 

Xenophon. Selections (for Schools). With Notes and Maps. By 

J. S. Phillpotts, B.CX. Fourth Edition. Extra fcap. 8vo. 3/. 6d. 



Anabasis, Book I. Edited for the use of Junior Classes 

and Private Students. With Introduction, Notes, etc. By J. Marshall, M.A., 
Rector of the Royal High School, Edinburgh. Extra fcap. 8vo. 2s. 6d. 

Anabasis, Book II. With Notes and Map. By C. S. 

Jerram, M.A. Extra fcap. 8to. a/. 

Cyropaedia, Book I. With Introduction and Notes by C. Bigg, 

D.D. Extra fcap. 8vo. %s. 

Cyropaedia, Books IV and V. With Introduction and 

Notes by C. Bigg, D.D. Extra fcap. 8to. ax. 6</. 



Aristotle^ s Politics. With an Introduction, Es.says, and Notes. 

By W. L. Newman, M.A., Fellow of Balliol Collie. Vols. I and II. 
Medium 8vo. aSj. 

Aristotelian Studies. I. On the Structure of the Seventh 

Book of the Nicomachean Ethics. By J.C. Wilson, M.A. 8vo. stiff, 5/. 

Aristotelis Ethica Nicotnachea, ex recensione Immanuelis 

Bekkeii. Crown 8vo. 51. 

Demosthenes and Aeschines. The Orations of Demosthenes 

and i&chines on the Crown. With Introductory Essays and Notes. By 
G. A. Simcox, M.A., and W. H. Simcox, M.A. 1873. 8vo. lax. 

Head [Barclay F.). Historia Numortim : A Manual of Greek 

Numismaiics, Royal 8yo. half-bound, a/, ax. 

Hicks (E. L.y M.A.). A Manual of Greek Historical Inscrip- 

twns, DemySvo. lor. 6</. 

Homer. Odyssey, Books I-XII. Edited with English Notes, 

Appendices, etc. By W. W. Merry, D.D., and the late James Riddell, M.A. 
1886. Second Edition. Demy 8vo. idr. 

Homer. A Grammar of tlte Homeric Dialect. By D. B. Monro, 

M.A. Demy Svo. lox. 6</. 

Polybius. Selections from Polybius. Edited by J. L. Strachan- 

Davidson, M.A., Fellow and Tutor of Balliol College. With three Maps. 
Medium 8to. buckram, au. Just PublisJud. 

Sophocles. The Plays and Fragments. With English Notes 

and Introductions, by Lewis Campbell, M.A. a vols. 1 

Vol. I. Oedipus Tyrannus. Oedipus Coloneus. Antigone. 8yo. i6x. \ 

Vol. II. Ajax. Electra. Trachiniae. Philoctetes. Fragments. 8yo. i6x. 



30 CLARENDON PRESS, OXFORD. 



ry*. FBENcn and Italian. 

Brachefs Etymological Dictionary of the French Languag^e. 

Translated by G. W. Kitchin, D.D. Third Edition. Crown 8vo. 7j. 6^. 

Historical Grammar of the French Language. Trans- 
lated by G. W. Kitchin. D.D. P'ourth Edition. Extra fcap. 8vo. JJ. 6//. 

Works by QEOBGE SAINTSBUBT, M.A. 

Primer of French Literature. Extra fcap. 8vo. is. 
Short History of French Literature. Crown 8vo. ioj.6rf 
Specimens of French Literature^ from Villon to Hugo. Crown 

8vo. 9/. 



MASTERPIECES OF THE FRENCH DRAMA. 
Comeill^s Horace. Edited, with Introduction and Notes, by 

George Saintsbury, M.A. Extra fcap. 8vo. 2j. 6^. 

Molihe^s Les Prkieuses Ridicules. Edited, with Introduction 

and Notes, by Andrew Lang, M.A. Extra fcap. 8vo. \s. 6d. 

Racine^s Esther. Edited, with Introduction and Notes, by 

George Saintsbury, M.A. ^xtra fcap. 8vo. 2s. 

Beaumarchais* Le Bar bier de SiviUe. Edited, with Introduction 

and Notes, by Austin Dobson. Extra fcap. 8vo. 2j. 6</. 

Voltair^s M^ope. Edited, with Introduction and Notes, by 

George Saintsbury. Extra fcap. 8yo. cloth, is. 

Mussefs On ne badine pas avec V Amour y and Fantasio. Edited, 

with Prolegomena, Notes, etc., by Walter Herries Pollock. Extra fcap. 
8yo. aj. 

The above six Plays may be had in ornamental case, and bound 
in Imitation Parchment, price i aj. 6^. 



Perraulis Popular Tales. Edited from the Original Editions, 

with Introduction, etc., by Andrew Lang, M.A. Small 4to. Hand-made 
paper, vellum back, gilt top, 15J. 

Sainte-Beuve. Selections from the Causeries du Lundi. Edited 

by George Saintsbury, M.A. Extra fcap. 8vo. 2s. 

Quinefs Lettres d sa Mire. Selected and edited by George 

Saintsbury, M.A. Extra fcap. 8vo. a.r. 

Gautier, Thiophile. Scenes of Travel. Selected and Edited 

by Greorge Saintsbury, M.A. Extra fcap. 8vo. %s. 

VAloquence de la Chaire et de la Tribune Franqaises, Edited 

by Paul Blouet, B.A. Vol. I. Sacred Oratory. Extra fcap. 8vo. ax. 6</. 



CLARENDON PRESS, OXFORD, 31 



SSdited by OUSTAVIB MA880N. B.A. 

Comeiiys Cinna. With Notes, Glossary, etc. Extra fcap. 8vo. 

chthf ax. Stiff covers, is. 6d, 

Louis XIV and his Contemporaries ; as described in Extracts 

from the best Memoirs of the Seventeenth Century. With English Notes, 
Genealogical Tables, &c. Extra fcap. 8vo. is. 6d,' 

Maistre, Xavier de. Voyage autotir de ma Chdmbre. Ourika, 

by Madame de Duras; Le Vieux Taillenr, by MM. Erckmann^Chatrian ; 
Lia Veill^e de Vinccnnes, by Alfred de Vigny ; Les Jnmeaux de THdtel 
ComtxWtyhy Edmond About : Mesaventures d*un ^colier, hy Rodolphe Topfftr. 
Third Edition, Revised and Corrected. Extra fcap. 8vo. ax. 6(/. 

Voyage autour de ma Chambre, Limp, u. 6^. 

Molikr^s Les Fourberies de Scapin, and Racine's Athalie, 

With Voltaire's Life of Moliire. Extra fcap. Svo. 2s. 6d. 

Molihre's Les Fourberies de Scapin. With Voltaire's Life of 

Moli^. Extra fcap. 8vo. stiff coveis, is. 6d. 

Moliire^s Les Femmes Savantes. With Notes, Glossary, etc. 

Extra fcap. Svo. cloth, ax. Stiff covers, ix. 6</. 

Racine^s Andromaque^ and CorneilUs Le Menteur. With 

Louis Racine's Life of his Father. Extra fcap. Svo. ax. 6d, 

Regnard*s Le ^otieur, and Brueys and Palaprafs Le Grondeur, 

Extra fcap. Svo. ax. 6d. 

Stvigni^ Madame de^ and her chief Contemporaries^ Selections 

from the Correspondence of. Intended more especially for Girls' Scliools. 
Extra fcap. Svo. 3x. 

Dante. Selections from the Inferno. With Introduction and 

Notes. By H. B. Cotterill, B.A. Extra fcap. Svo. 4X. (k/. 

Tasso. La Gerusalemme Liberata. Cantos i, ii. With In- 
troduction and Notes. By the same Editor. Extra fcap. Svo. ax. 6d. 

V. GERMAN. 
Scherer ( WX A History of German Literature. Translated 

from tne Tnird German Edition by Mrs. F. Conybeare. Edited by F. Max 
Miiller. a vols. Svo. aix. 

Max Miiller* The German Classics, from the Fourth to the 

Nineteenth Century. With Biographical Notices, Translations into Modem 1 

German, and Notes. By F. Max Miiller, M.A. A New Edition, Revised, • 

Enlarged, and Adapted to Wilhelm Scherer's ' History of German Literature,' ' 
by F. Lichtenstein. a vols, crown Svo. aix. 



32 CLARENDON PRESS, OXFORD. 



GERMAN COURSE, ByHBBMANlT IiANOSS. 

The Germans at Hotne; a Practical Introduction to German 

Conversation, with an Appendix containing the Essentials of Gennan Grammar. 
Third Edition. 8vo. 2s, 6d, 

The German Manual; a German Grammar, Reading Book, 

and a Handbook of German Conversation. 8vo. 7^. 6^. 

Grammar of the German Language, 8vo. y. 6d. 

German Composition ; A Theoretical and Practical Guide to 

the Art of Translating English Prose into German. Ed. a. 8vo. \s. 6d, 

German Spelling; A Synopsis of the Changes which it has 

undergone through the Government Regulations of i88o. Paper covers, 6d. 



Lessing^s Laokoon. With Introduction, English Notes, etc. 

By A. Hamann, Phil. Doc, M.A. Extra fcap. 8vo. 4s, 6d. 

Schiller^s WiUielm Tell. Translated into English Verse by 

E. Massie, M.A. Extra fcap. 8vo. 51. 

GERMAN CLASSICS. 

With Biographical, Historical, and Critical Introductions^ Arguments 
(to the Dramas), and Complete Commentaries. 

Edited by C. A. BITCHHISIM, Phil. Doo. Professor in King's 

OoUege, Iiondon. 

Lessing: 

(a) Nathan der Weise. A Dramatic Poem. 4s. 6d, 
\b) Minna von Barnhelm, A Comedy. 35. 6d, 

Goethe : 

(a) Egmont, A Tragedy. 3J. 

(d) Iphigenie auf Tauris. A Drama. 3J. 

Schiller : 

(a) Wilhelm Tell. A Drama. Large Edition. With a Map. 3^. td. 
\b) Wilhelm Tell. School Edition. With a Map. zr. 
\c) Historische Skizzen. With a Map. is. 6d. 

Heine: 

fa) Prosa : being Selections from his Prose Writings. \s. 6^. 
\b) Harzreise. Cloth, 2s. 6d, ; paper covers, is. 6d. 

Modem German Reader. A Graduated Collection of Ex- 
tracts from Modern German Authors : — 
Part I. Prose Extracts. With English Notes, a Grammatical Appendix, and 

a Complete Vocabulary. Fourth Edition. a5.6</. 
Part II. Extracts in Prose and Poetry. With English Notes and an Index. 
Second Edition. 2s, 6d, 



ii 



CLARENDON PRESS, OXFORD. 33 



Becker (the Historian) : 

Friedrich der Grosse, Edi*^ed, with Notes, an Historical Introduc- 
tion, and a Map. 3^. td. Just Published. 



Niebuhr : 

Griechische Heroen-Geschichten (Tales of Greek Heroes). Edited, 
with English Notes and a Vocabulary, by Emma S. Budiheim. Second. 
Revised Edition, clothj is. 



A Middle High German Primer, With Grammar, Notes, 

and Glossary. By Joseph Wright, Ph. D. Extra fcap. 8vo. jj. 6i/. 
VI. MATHEMATICS, PHYSICAL SCIBNCE, &o. 

By I1ISWI8 HENSIiB7, M.A. 

Figures made Easy : a first Arithmetic Book. Crown 8vo. 6d. 
Answers to the Examples in Figures made Easy^ together 

with two thousand additional Examples, with Answers. Crown 8vo. u. 

The Scholar^ s Arithmetic. Crown 8vo. %s. 6d, 

Answers to the Examples in the Scholar s Arithmetic, is. 6d. 

The Scholar^ 5 Algebra. Crown 8vo. %s. 6d. 



Aldis ( W. S.y M.A.). A Text-Book of Algebra: with Answers 

to tfu Examples, Crown 8vo. 7^. 6^. 

Baynes {R, E.yM.A,). Lessons on Thermodynamics. 1878. 

Crown 8vo. yj. td. 

Chambers {G. /^, F.R.A.S.). A Handbook of Descriptive 

Astronomy, Third Edition. 1877. Demy 8vo. 28^. 

Clarke {Col. A, R.,C.B.yR.E.). Geodesy. 1880. 8vo. i%s.(^d. 

Cremona {Luigi). Elements of Projective Geometry. Trans- 
lated by c. Leudesdorf, M.A. 8vo. I2s,6d, 

Donkin. Acoustics. Second Edition. Crown 8vo. Ts. 6d. 

Euclid Revised. Containing the Essentials of the Elements 

of Plane Geometry as given by Euclid in his first Six Books. Edited by 
R. C. J. Nixon, M.A. Crown 8vo. •js. 6d. 

Bold separately m follows, 

Book I. I J. Books 1, II. u. 6d. 

Books I-IV. is.ed. Books V, VI, 3/. 

Euclid. — Geometry in Space, Containing parts of Euclid's 

Eleventh and Twelfth Books. By the same Editor. Crown 8vo. ^. td, 

D 



34 CLARENDON PRESS, OXFORD. 

Galton {Douglas, C.B., F.R.S.), TIu Construction of Healthy 

Dwellings, Demy 8vo. lor. 6^. 

Hamilton {Sir R. G. C), and y. BalL Book-keeping. New 

* and enlarged Edition. Extra fcap. 8vo. limp cloth, aj. 

Ruled Exercise books adapted to the above may be had, price 2s. 

Har court (A. G, Vernon, M.A.), and H, G, Madan^ M.A, 

Exercises in Practical Chemistry. Vol. I. Elementary Exercises. Fourth 
Edition. Crown 8vo. los. 6</. 

Maclaren {Archibald). A System of Physical Education : 

Theoretical and Practical. Extra fcap. 8vo. 7j. 6</. 

Madan {H. G., M.A,). Tables of Qualitative Analysis. 

Large 4to. paper, \s, 6d, 

Maxwell {y. Clerk, M.A., F.R.S.). A Treatise on Electricity 

and Magnetism, Second Edition, a vols. Demy 8vo. i/. \\s. 6d. 

An Elementary Treatise on Electricity. Edited by 

William Gamett, M.A. Demy 8vo. 7j. (td. 

Minchin {G. M., M.A.). A Treatise on Statics with Applica- 
tions to Physics. Third Edition, Corrected and Enlarged. Vol. I. Equili- 
brium ofCoplanar Forces, 8vo. 9/. Vol. II. Statics, 8vo. i6j. 

Uniplanar Kinematics of Solids and Fluids. Crown 



8vo. 7j. 6d, 

Phillips {yohn, M.A., F.R.S.). Geology of Oxford and the 

Valley of the Thames, 1871. 8vo. au. 



Vesuvius. 1869. Crown 8 vo. ios.6d. 



Prestwich {yoseph, M.A., F.R.S.). Geology, Chemical,Physical, 

and Stratigraphical, In two Volumes. 

Vol. I. Chemical and Physical. Royal 8vo. a5x. 

Vol. II. Stratigraphical and Physical. With a new Geographical Map of 
Europe. Royal 8vo. 36J. Just published. 

Rolleston {George^ M,D., F,R,S.). Forms of Animal Life. 

A Manual of Comparative Anatomy, with descriptions of selected types. 
Second Edition. Revised and enlarged by W. Hatchett Jackson, M.A. 
Medium, 8vo. cloth extra, 1/. \^s. 

Smyth. A Cycle of Celestial Objects. Observed, Reduced, 

and Discussed by Admiral W. H. Smyth, R.N. Revised, condensed, and 
greatly enlarged by G, F. Chambers, F.R.A.S. 1881. 8vo. 12/. 

Stewart {Balfour, LL.D,, F.R.S.). An Elementary Treatise 

on Heaif with numerous Woodcuts and Diagrams. Fifth Edition. Extra 
fcap. 8vo. 7x. 6</. 



CLARENDON PRESS, OXFORD, 35 



Vemon-Harcourt {L. F,^ M.A.). A Treatise on Rivers and 

Canals, relating to the Control and Improvement of Rivers, and the Design, 
Construction, and Development of Canals, a vols. (Vol. I, Text. Vol. II, 
Plates.) 8vo. 21s, 

Harbours and Docks ; their Physical Features, History, 



Construction, Equipment, and Maintenance ; with Statistics as to their Com- 
mercial Development, a vols. 8vo. 35J. 

Walker {James , M.A.), The Theory of a Physical Balance, 

8vo. stiff cover, y. 6d. 

Watson {H, W.y M.A.), A Treatise on the Kinetic Theory 

of Gcues. 1876. 8vo. $s.6d. 

Watson {H. W., D. Sc, F.R.S.), and S. H. Burbury, M.A. 

I. A Treatise on the Application of Generalised Coordinates to the Kinetics of 

a Material System, 1879. 8vo. 6j. 

II. The McUhematical Theory of Electricity and Magnetism, Vol. I . Electro- 
statics. 8vo. I Of. dd. 

Williamson (A. W., Phil. Doc.^ F.R.S.). Chemistry for 

Students, A new Edition, with Solutions. 1873. ^xtra fcap. 8vo. 8f. 6</. 



VII. HISTORY. 

Bluntschli {J. K.). The Theory of the State. By J. K. 

Bluntschli, late Professor of Political Sciences in the University of HeideU 
berg. Authorised English Translation from the Sixth German Edition. 
Demy 8vo. half bound, i ax. td, 

Finlay {George^ LL.D,). A History of Greece from its Con- 
quest by the Romans to the present time, B.C. 146 to a.d. 1864. A new 
Edition, revised throughout, and in part re-written, with considerable ad- 
ditions, by the Author, and edited by H. F. Tozer, M.A. 7 vols. 8vo. 3/. lOf. 

Fortescue (Sir John, Kt,). The Governance of England: 

otherwise called The Difference between an Absolute and a Limited Mon- 
archy. A Revised Text. Edited, with Introduction, Notes, and Appendices, 
by Charles Plummer, M.A. 8vo. half bound, lax. 6</. 

Freeman {E.A., D,C.L.), A Short History of the Norman 

Conquest of England, Second Edition. Extra fcap. 8vo. 2s,6d, 

George (H. B. , M. A .). Genealogical Tables illustrative of Modern 

History. Third Edition, Revised and Enlarged. Small 4to. 1 2s, 

Hodgkin {TX Italy and Iter Invaders. Illustrated with 

Plates and Maps. Vols. I— IV, a.d. 376-553. 8vo. 3/. 8j. 

D 2 



36 CLARENDON PRESS, OXFORD, 

Hughes (Alfred), Geography for Schools, With Diagrams. 

Part I. Practical Geography. Crown 8vo. 2j. 6d. Just Published, 
Part II. General Geography. In preparation. 

Kitchin (G. W., D.D,), A History of France, With numerous 

Maps, Plans, and Tables. In Three Volumes. Second Edition, Crown 8vo. 
each I Of. 6</. 

Vol. I. Down to the Year 1453. 

Vol.11. From 1453-1624. Vol. III. From 1624-1793. 

Lucas (C. -P.). Introduction to a Historical Geography of tJte 

British Colonies. With Eight Maps. Crown 8vo. 4J. 6^. 

Payne {E, J,, M,A,), A History of the United States of 

America, In the Press. 

Ranke {L. von), A History of England, principally in the 

Seventeenth Century. Translated by Resident Members of the University of 
Oxford, under the superintendence of G» W. Kitchin, D.D., and C. W. Boase, 
M.A. 1875. 6 vols. 8vo. 3/. 3J. 

Rawlinson {George , M,A,), A Manual of Ancient History. 

Second Edition. Demy8vo. 14J. 

Ricardo. Letters of David Ricardo to Thomas Robert Mcdthus 

(1810-1823). Edited by James Bonar, M. A. Demy 8vo. i oj. 6</. 

Rogers {J, E. Thorold, M,A,), The First Nine Years of the 

Bank of En^and. 8vo. 8;. td. 

Select Charters and other Illustrations of English Constitutional 

History, from the Earliest Times to the Reign of Edward I. Arranged and 
edited by W. Stubbs, D.D. Fifth Edition. 1883. Crown 8vo. 8j. 6d. 

Stubbs ( W,^ D,D,), The Constitutional History of Efigland^ 

in its Origin and Development. Library Edition. 3 vols, demy 8vo. 2/. 8j. 
Also in 3 vols, crown 8vo. price \is. each. 

Seventeen Lectures on the Study of Medieval and 



Modem History^ &c., delivered at Oxford 186 7-1 884. Crown 8vo. %s, 6d, 

Wellesley, A Selection from the Despatches^ Treaties^ and 

other Papers of the Marquess Wellesley. K.G., during his Government 
of India. Edited by S. J. Owen, M.A. 1877. 8vo. 1/. 4J. 

Wellington, A Seleciioft from the Despatches, Treaties, and 

other Papers relating to India of Field-Marshal the Duke of Wellington, K.G. 
Edited by S. J. Owen, M.A. 1880. 8vo. 24^. 

A History of British India. By S. J. Owen, M.A., Ijleader 

in Indian History in the University of Oxford. In preparation. 



CLARENDON PRESS, OXFORD. 37 



VIII. LAW. 

Alberici Gentilis, LCD., I.C., De lure Belli Libri Tres. 

EdidttT. £. Holland, LCD. 1877. Small 410. half morocco, au. 

Anson {^Sir William J?., BarL^ D.CL,). Prvtciples of the 

English Law of Contract ^ and of Agency in its Relation to Contract. Fifth 
Edition. Demy 8vo. ioj. 6</. 

Law and Custom of the Constitution* Part I. Parlia- 



ment. Demy Svo. lor. 6^. 

Bentham {Jeremy), An Introduction to the Principles of 

Morals and Legislation. Crown Svo. 6j. dd, 

Digby (Kenelm E., M,A,), An Introduction to the History of 

the Law of Real Property. Third Edition. Demy Svo. ioj. 6</. 

Gaii Institutionum Juris Civilis Commentarii Quattuor ; or, 

Elements of Roman Law by Gains. With a Translation and Commentary 
by Edward Poste. M.A. Second Edition. 1875. Svo. i8j. 

Hall ( W. E,^ MA.). International Law. Second Ed. Svo. ai j. 
Holland {T. E.^ D.C.L.). TJu Elements of Jurisprudence. 

Fourth Edition. Demy Svo. \os. 6^. 

The European Concert in the Eastern Question^ a Col- 
lection of Treaties and other Pnblic Acts. Edited, with Introductions and 
Notes, by Thomas Erskine Holland, D.CL. Svo. i %s. 6d, 

Imperatoris lustiniani Institutionum Libri Quattuor; with 

Introductions, Commentary, Excursus and Translation. By J. B. Moyle, B.C.L., 
M.A. a vols. Demy Svo. au. 

Justinian^ The Institutes of edited as a recension of the 

Institutes of Gaius, by Thomas Erskine Holland, D.CL. Second Edition, 
1 88 1 . Extra fcap .Svo. 5 j. 

JustiniaUy Select Titles from the Digest of By T. E. Holland, 

D.CL., and C L. Shad well, B.CL. Svo. \\s. 

Also sold in Farts, in paper covers, as follows : — 

Part I. Introductory Titles, as. 6rf. Part 11. Family Law. if. 

Part III. Property Law. as. 6rf. Part IV. Law of Obligations (No. 1). 3s. W. 

Part IV. Law of Obligations (No. a). 4s. 6J. 

Lex Aquilia. The Roman Law of Damage to Property : 

being a Commentar)' on the Title of the Digest • Ad Legem Aquiliam * (ix. 2). 
With an Introduction to the Study of the Corpus luris Civilis. By Erwin 
Grueber, Dr. Jur., M.A. Demy Svo. lor. 6</. 



38 CLARENDON PRESS, OXFORD. 

Markby ( fF., D.C.L,). Elements of Law considered with refer- 
ence to Principlesof General Jurisprudence. Third Edition. DemySvo. 12s, 6d. 

Stokes ( Whitley, D.C.L,). The Anglo-Indian Codes. 

Vol. I. Sttbstantive Law, 8vo. .^or. 
Vol. II. Adjective Law, In the Press. 

Twiss (Sir Travers, D.CX.). The Law of Nations considered 

as Independent Political Communities. 

Part I. On the Rights and Duties of Nations in time of Peace. A new Edition, 
Revised and Enlarged. 1884. DemySvo. 15J. 

Part II. On the Rights and Duties of Nations in Time of War. Second Edition, 
Revised. 1875. Demy 8vo. au. 



IX. HEKTAL AKD MOBAL PHILOSOPHY, &o. 

BacofCs Novum Organum. Edited, with English Notes, by 

G. W. Kitchin, D.D. 1855. 8vo. 9J. 6rf. 

Translated by G. W. Kitchin, D.D. 1855. 8vo. 9J. 6d. 

Berkeley, The Works of George Berkeley, D,D., formerly 

Bishop of Cloyne ; including many of his writings hitherto unpublished. 
With Prefaces, Annotations, and an Account of his Life and Philosophy, 
by Alexander Campbell Eraser, M.A. 4 vols. 1871. 8vo. 2/. i8j. 

The Life, Letters, &c. i vol. i6j. 

Berkeley, Selections from. With an Introduction and Notes. 

For the use of Students in the Universities. By Alexander Campbell Fraser, 
LL.D. Third Edition. Crown 8vo. 71. 6</. 

Fowler ( T., D.D.), The Elements of Deductive Logic, designed 

mainly for the use of Junior Students in the Universities. Ninth Edition, 
with a Collection of Examples. Extra fcap. 8vo. 3/. 6^/. 

The Elements of Inductive Logic, designed mainly for 

the use of Students in the Universities. Fourth Edition. Extra fcap. 8vo. 6j. 

- and Wilson {J. M., B.D.). The Principles of Morals 



(Introductory Chapters). 8vo. boards, ^s. 6d, 

The Principles of Morals. Part II. (Being the Body 



of the Work.) 8vo. \os,(id. 

Edited by T. FO'WIiEB, D.D. 

Bacon. Novum Organum. With Introduction, Notes, &c. 

1878. 8vo. 14J. 

Locke's Conduct of the Understanding. Second Edition. 

Extra fcap. 8vo. 2s,